Human acute myeloid leukemia cells MOLM13, OCI-AML2, NB4, HL-60, and U937 were kindly provided by Professor Jie Jin from the Institute of Hematology at the First Affiliated Hospital of Zhejiang University. KG-1α cells were obtained from the American Type Culture Collection. MOLM13, OCI-AML2, and KG-1α cells were maintained in Iscove’s Modified Dulbecco Medium (IMDM) supplemented with 10% fetal bovine serum (FBS). U937, NB4, and HL-60 cells were incubated in PRMI 1640 medium. All cells were cultured at 37°C in a humidified atmosphere containing 5% CO₂.

In addition, mononuclear cells were isolated from the bone marrow samples collected from patients diagnosed with AML by Ficoll–Hypaque (Sigma-Aldrich, United States) density gradient centrifugation. Anticoagulated whole blood or marrow was mixed with an equal volume of PBS containing 2% FBS, carefully layered over Ficoll solution, and centrifuged at 400 × g for 30 min. The buffy coat layer was then collected, washed twice with PBS (300 × g, 10 min each), and finally resuspended in RPMI-1640 medium supplemented with 10% FBS. Cell viability (>95%) was confirmed by trypan blue exclusion. Written informed consent was obtained from all participating patients. The study protocol was approved by the Ethics Committee of the First Affiliated Hospital of Ningbo University, China (NBU-2023-081).

Algae-derived fucoxanthin (99.47% purity) was purchased from MedChem Express (Monmouth Junction, NJ, United States).

Fucoxanthin (1.56–200 μM; 2-fold dilutions) was tested on MOLM13 cells (5 × 10⁴/well) for 24 h. The IC₅₀ value was calculated using a four-parameter logistic model, and 25 μM (1.5× IC₅₀) was selected as the optimal dose based on preliminary kinetics data. MOLM13, NB4, HL-60, OCI-AML2, KG-1α, and U937 cells were seeded in 96-well plates at a density of 1 × 10⁴ cells per well. The cells were then treated with 6.25, 12.5 and 25 μM natural carotenoids (fucoxanthin, astaxanthin, and lycopene) for 24 h. The mononuclear cells isolated from bone marrow were treated with 25 μM fucoxanthin for 24 or 48 h. Subsequently, the cell counting kit-8 solution (10 μL) (NCM Biotech, Shanghai, China) was added to each well and the cells were incubated for an additional 2 h, and the spectrophotometric absorbance at 450 nm was measured. All experiments were performed in triplicate. The results were expressed as a percentage of the control (Zhang et al., 2025).

Apoptosis induction was assessed using an annexin V-APC/7-AAD apoptosis kit (Liankebio, Hangzhou, China). After treatment with fucoxanthin (6.25, 12.5 and 25 μM) for 24 h, cells were washed twice with cold phosphate-buffered saline (PBS), suspended in the binding buffer, incubated with annexin V-APC and 7-AAD for 15 min, then analyzed by flow cytometry using a FACScan flow cytometer (Becton Dickinson, San Diego, CA, United States) (Wu et al., 2022).

Cell cycle arrest analysis was performed using a propidium iodide DNA staining kit (Liankebio, Hangzhou, China). After treatment with fucoxanthin (6.25, 12.5 and 25 μM) for 24 h, cells were harvested, fixed with 70% ethanol (1 mL per 1 × 10⁶ cells) at −20°C overnight, and hydrated with PBS at room temperature for 15 min. The DNA staining solution (1 mL) was then added, and the cells were incubated at room temperature in the dark for 30 min before analysis by flow cytometry using a FACScan flow cytometer (Becton Dickinson, San Diego, CA, United States) (Li et al., 2022).

The fluorescent d-glucose analog 2-(N-(7-nitrobenz-2-oxa-1,3-diazol-4-yl)amino)-2-deoxyglucose (2-NBDG) (MedChemExpress, Monmouth Junction, NJ, United States) was employed as a fluorescent indicator to assess glucose uptake. MOLM13 cells were treated with fucoxanthin (6.25, 12.5 and 25 μM) for 24 h, harvested and washed with PBS, starved in glucose-free medium for 1 h, then incubated with 2-NBDG (100 μM) at 37°C under 5% CO₂ for 30 min. The cells were then harvested by centrifugation, washed with cold PBS twice, and the 2-NBDG signal was analyzed using a FACScan flow cytometer (Becton Dickinson, San Diego, CA, United States) (Qiu et al., 2025).

MOLM13 cells were seeded in 6-well plates at a density of 1 × 10⁶ cells per mL, treated with fucoxanthin (6.25, 12.5 and 25 μM) for 24 h, and washed twice with PBS. The ATP content was measured using an ATP assay kit (Beyotime, Shanghai, China) following the manufacturer’s instructions. Specifically, cells were lysed with ice-cold lysis buffer (200 μL per 1 × 10⁶ cells, 6-well plate equivalent), vortexed for 10 s, incubated on ice for 10 min, and centrifuged at 12,000 × g and 4°C for 5 min; the supernatants were collected for analysis. ATP standards (0.01–10 μM) were prepared by serial dilution in lysis buffer, and the working solution was created by mixing the ATP detection reagent with the assay diluent (1:9 v/v). For detection, the working solution (100 μL) was added to luminometer plates, incubated at room temperature for 5 min to permit background ATP hydrolysis, followed by the injection of the sample or standard (20 μL). Luminescence (RLU) was then measured within 2 s (Wang et al., 2022).

AKT kinase activity was measured using a commercial colorimetric assay kit (Genmed Scientifics Inc., United States) with minor modifications. Cells cultured in 6-well plates were lysed in ice-cold lysis buffer (100 μL), vortexed for 15 s, and incubated on ice for 30 min. The lysates were centrifuged (12,000 × g, 4°C, 10 min), and the supernatants were collected. The total protein concentration was determined via BCA assay (Easy II Quantitative Kit, TransGen Biotech, Beijing, China). For the reaction, the total protein (50 μg) was combined with the assay buffer to attain a final volume of 200 μL, and the mixture was incubated at 30°C for 40 min. The enzyme activity was determined by monitoring NADH oxidation at 340 nm, and one unit of AKT activity was defined as the oxidation of 1 μmol NADH/min/mg protein at pH 7.5°C and 30°C (Fang et al., 2021).

The crystal structure of the AKT protein (PDB:4GV1) was downloaded from the PDB database (http://www.rcsb.org/). The 3D structure of fucoxanthin was obtained from the PUBCHEM database (https://pubchem.ncbi.nlm.nih.gov/), and energy minimization was performed using the MMFF94 force field. The AutoDock Vina 1.1.2 software was used for molecular docking. The receptor protein was processed using PyMol 2.5 to remove water molecules, salt particles, and small molecules. The center of the box (25 × 25 × 25 Å3) was set to the centroid of the ligand in the original crystal structure. All processed small molecules and the receptor protein were, using ADFRsuite 1.03, converted into the PDBQT format required for docking. During docking, the exhaustiveness of the global search was set to 32, and the default settings were used for other parameters. The docked conformation with the highest score was considered the binding conformation, and the docking results were visualized using PyMol 2.5 (Wu et al., 2021).

Molecular dynamics simulations were performed using the AMBER 18 software. The system first underwent energy minimization using a combination of 2,500 steps of steepest descent and 2,500 steps of conjugate gradient method, and it was then heated at a constant rate over 200 ps from an initial temperature of 0 K to the desired simulation temperature of 298.15 K while the volume remained fixed. Once the system reached 298.15 K, a 500 ps simulation was performed in the NVT ensemble to allow the solvent molecules to distribute evenly within the solvent box. Afterwards, a 500 ps equilibrium simulation of the entire system was performed under NPT (constant temperature and pressure) conditions.

Two complex systems were then simulated under periodic boundary conditions for a total of 30 ns each using the NPT ensemble. During these simulations, a non-bonded cutoff distance of 10 Å was applied to determine the interactions between atoms. The particle mesh Ewald (PME) method was employed to calculate long-range electrostatic interactions. The SHAKE algorithm was utilized to constrain the length of hydrogen bonds in the system. To control the temperature, Langevin dynamics was applied with a collision frequency of 2 ps⁻¹. The system pressure was maintained at 1 atm, and the integration time step was 2 fs. The trajectories of the simulations were saved every 10 ps. The binding free energy between the protein and the ligand was calculated using the MM/GBSA method as follows: ∆ G b i n d = ∆ G c o m p l e x − ∆ G r e c e p t o r + ∆ G l i g a n d = ∆ E int e r n a l + ∆ E V D W + ∆ E e l e c + ∆ G G B + ∆ G S A where ΔG is the Gibbs free energy, and ΔE _internal, ΔE _VDW, and ΔE _elec represent the internal energy, van der Waals interactions, and electrostatic interactions, respectively. The internal energy involves the bond energy (E _bond), the angle energy (E _angle), and the torsional energy (E _torsion). The solvation-free energy is collectively referred to as ΔG _GB and ΔG _SA , where G _GB represents the polar solvation free energy and G _SA represents the non-polar solvation free energy. The generalized Born (GB) model developed by Nguyen et al. (igb = 8) was used to calculate ΔG _GB. The non-polar solvation free energy (ΔG _SA) was calculated as the product of the surface tension and the solvent-accessible surface area (SASA), i.e., ΔG _SA = 0.0072 × ΔSASA (Needham et al., 2016; Salomon-Ferrer et al., 2013).

The AKT overexpression plasmid pEX-3(pGCMV/MCS/Neo)-AKT and the empty vector plasmid pEX-3(pGCMV/MCS/Neo) were purchased from GenePharma (Shanghai, China) and transfected into HEK-293T cells using Lipofectamine^® 3,000 (Thermo Fisher Scientific, Waltham, MA, United States) (Ma et al., 2014). Twenty-four hours after the transfection, the cells were treated with fucoxanthin and incubated for an additional 24 h. The cells were then collected, and the expression levels of poly ADP ribose Polymerase (PARP) and GLUT1 were measured by Western blot. In addition, the AKT kinase activity and the ATP levels were quantified using the corresponding assay kits.

The total protein was extracted using a total protein extraction kit (Beyotime, Shanghai, China) following the manufacturer’s instructions. The protein concentration was determined using the BCA Kit (TransGen, Beijing, China). To begin the Western blot analysis, equal amounts of proteins (30 μg) from the whole cell lysates were separated using 10% sodium dodecyl sulfate-polyacrylamide gel (SDS-PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes. The membranes were blocked with 5% skim milk for 2 h at room temperature, incubated at 4°C with the primary antibodies overnight, then incubated with the appropriate secondary antibodies at room temperature for 1 h. The immunoreactive bands were visualized by electrochemiluminescence using NcmECL Ultra-HRP Substrate (New Cell & Molecular Biotech Co., Ltd., Suzhou, China). The band intensity was quantified using ImageJ with β-actin as the control. The primary antibodies included Bcl-2 (1:1000), Bax (1:1000), GLUT1 (1:1000), PARP (1:1000), CDK4 (1:1000), CDK6 (1:1000), and β-actin (1:2000) (Cell Signaling Technology, Danvers, MA, United States). The secondary antibodies included horseradish peroxidase (HRP)-conjugated mouse anti-rabbit IgG (1:5000) and goat anti-mouse IgG (1:5000) antibodies (Cell Signaling Technology, Danvers, MA, United States) (Niu et al., 2018).

Data were presented as the mean ± standard deviation (SD). Statistical analyses were conducted using SPSS 16.0 (SPSS Inc., Chicago, IL, United States). The statistical significance was assessed using one-way ANOVA with the Tukey multiple comparison test. Differences were considered statistically significant when P < 0.05.

Among the three tested natural carotenoids, astaxanthin and lycopene did not have a noticeable effect on MOLM13 cells, whereas fucoxanthin significantly inhibited the growth of MOLM13 cells (Figures 1A–C). The survival of MOLM13 cells reduced by 63.6% when they were treated with 25 μM fucoxanthin. Figure 1D shows that among the AML cell lines in the growth test, MOLM13 cells were inhibited by fucoxanthin most effectively, followed by KG-1α and NB4 cells, while no apparent effect was detected for OCI-AML2 or U937 cells (P > 0.05). In addition, when bone marrow-derived primary cells were treated with 25 μM fucoxanthin, the cell survival decreased by 72.9% after 24 h and by 85.2% after 48 h (Figure 1E).

After MOLM13 cells were treated with 25 μM fucoxanthin for 24 h, compared to the control group, the proportion of cells in the G0/G1 phase increased by 118.72%, whereas the proportion of cells in the S phase decreased by 63.8% (Figure 2A). The expression levels of CDK4 and CDK6 proteins, which are involved in cell cycle regulation, were investigated to examine the mechanism underlying the observed cell cycle arrest. Interestingly, while the treatment with 25 μM fucoxanthin for 24 h decreased the CDK4 expression in MOLM13 cells by 48.6%, the CDK6 expression was not affected (Figure 2B). Thus, it could be inferred that fucoxanthin induced cell cycle arrest in MOLM13 cells at the G0/G1 phase by inhibiting the expression of the CDK4 protein.

Cells may undergo apoptosis to trigger programmed death after prolonged cell cycle arrest. Indeed, the apoptosis of MOLM13 cells increased with rising concentration of fucoxanthin used in the incubation (Figure 3A). Interestingly, the treatment with 25 μM fucoxanthin, which raised the apoptosis rate of MOLM13 cells up to 38.3%, decreased the Bcl2 expression by 75.3% (P < 0.01) but did not affect the Bax expression (Figure 3B), and it increased the cleavage of the PARP protein remarkably (Figure 3C, P < 0.01). Therefore, fucoxanthin induced apoptosis in MOLM13 cells by enhancing the cleavage of the PARP protein and inhibiting the expression of the Bcl2 protein.

MOLM13 cells had higher levels of the AKT protein than other cell lines, and U937 and OCI-AML2 cells had the lowest levels of the AKT protein (Figure 4A). In addition, the AKT expression was significantly stronger in the bone marrow-derived primary cells than in the peripheral blood cells (Figure 4B). Molecular docking was used to assess the binding between fucoxanthin and AKT. Figure 4C illustrates that fucoxanthin is capable of specific binding with AKT, and the affinity is −7.5 kcal/mol. The active pocket on the AKT protein that envelops fucoxanthin features Phe-237, Phe-236, Asp-439, Phe-442, Phe-161, Lys-179, and Leu-295. Fucoxanthin engages Phe-237, Phe-236, Asp-439, Phe-442, and Phe-161 through hydrophobic interactions and Lys-179 and Leu-295 through hydrogen bonding. Molecular dynamics simulation shows that fucoxanthin has a binding energy of −30.38 kcal/mol to the AKT protein. The amino acids Lys-179 and Phe-236 are primarily involved in the binding, and they contribute −2.15 and −1.89 kcal/mol to the binding energy, respectively (Figure 4D). Fucoxanthin decreased the kinase activity of AKT in a concentration-dependent manner (Figure 4E), and the AKT kinase activity dropped by 57.9% (P < 0.01) after MOLM13 cells were treated with 25 μM fucoxanthin for 24 h.

The GLUT1 expression of MOLM13 cells dropped significantly after incubation with fucoxanthin at a higher concentration (Figure 5A), and their glucose absorption also declined with rising fucoxanthin concentration in a dose-dependent manner (Figure 5B). In addition, fucoxanthin also significantly reduced the cellular ATP content (Figure 5C). After MOLM13 cells were incubated with 25 μM fucoxanthin for 24 h, the GLUT1 expression decreased by 36.1% (P < 0.01), the glucose absorption dropped by 75.35% (P < 0.01), and the intracellular ATP fell by 71.7% (P < 0.01). Presumably, these results suggested that fucoxanthin may impair glucose metabolism in MOLM13 cells, possibly through inhibition of GLUT1-mediated glucose transport, as evidenced by reduced glucose uptake and subsequent ATP depletion.

A cell model with AKT overexpression was built to further examine the effects of fucoxanthin. The overexpression of the AKT protein increased the AKT kinase activity (P < 0.01, Figure 6A), upregulated GLUT1 expression (P < 0.05, Figures 6B,C), and increased the intracellular ATP (P < 0.01, Figure 6C). The AKT overexpression did not have any obvious effect on the cleavage of the PARP protein (Figure 6B). When the cell model was treated with 25 μM fucoxanthin for 24 h, the AKT kinase activity dropped by 12.6% (P > 0.05), the GLUT1 expression decreased by 34.1% (P < 0.05), although the ATP level changed very little. The activation of PARP cleavage induced by fucoxanthin was hindered after AKT overexpression (Figure 6B). The results suggested that fucoxanthin targeted the AKT protein to inhibit ATP production and activate PARP cleavage, thereby promoting cell cycle arrest and apoptosis (Figure 7).