Potencial antiproliferativo y antiinflamatorio del 2-cloro-3-((4-hidroxifenil)amino)naftaleno-1,4-diona mediante la modulación de las vías de señalización TNF-α/IL-8 y NF-κB en células cancerosas HCT116: un estudio exhaustivo in silico y in vitro.

Cancer is one of the main causes of mortality globally, according to the World Health Organization (WHO) [^[1]]. Colorectal cancer (CRC) is among the most common malignancies globally, and accumulating evidence indicates that persistent inflammation plays a critical role in colorectal tumor initiation, progression, and metastasis [^[2]]. Chronic inflammatory signaling promotes the formation of a pro-tumorigenic microenvironment by recruiting immune cells, activating oncogenic signaling pathways such as nuclear factor-kappa B (NF-κB) and signal transducer and activator of transcription 3 (STAT3), and stimulating the production of pro-inflammatory mediators, including tumor necrosis factor-alpha (TNF-α) and interleukins (IL-6, IL-8, IL-10, and IL-17) [^[3],^[4]]. Among these signaling pathways, NF-κB has emerged as a central regulator linking chronic inflammation to malignant transformation by controlling the expression of genes involved in cytokine production, cell survival, proliferation, angiogenesis, and metastasis [^[5]].

The five members of the NF-κB transcription factor family—NF-κB1 (p105/p50), NF-κB2 (p100/p52), p65 (RELA), RELB, and c-REL—control the expression of genes related to innate and adaptive immunity, inflammation, cell proliferation, apoptosis, and tissue remodeling [^[6]]. In resting cells, NF-κB dimers are sequestered in the cytoplasm through their association with inhibitor of NF-κB (IκB) proteins. Upon stimulation by pro-inflammatory cytokines such as TNF-α, microbial products, or oxidative stress, the IκB kinase (IKK) complex phosphorylates IκB proteins, triggering their ubiquitination and subsequent proteasomal degradation. This process releases NF-κB, allowing its translocation into the nucleus, where it activates the transcription of numerous target genes involved in inflammation, cell survival, proliferation, and immune responses [^[7]]. Among these targets, TNF-α and IL-8 are particularly important because they amplify inflammatory signaling, promote recruitment of immune cells, and contribute to the establishment of a tumor-promoting microenvironment that facilitates cancer growth and progression [^[8]]. Previous studies have also shown that persistent activation of this NF-κB-centered inflammatory network is implicated in sustained mucosal inflammation, resistance to apoptosis, and enhanced malignant behavior in CRC [^[9]]. Furthermore, NF-κB supports tumor growth, vascularization, invasion, and metastasis via controlling genes that promote angiogenesis and invasion, such as VEGF, IL-8, COX-2, and MMP-9, and cell-cycle progression, such as Cyclin D1 [^[5],^[10],^[11]]. Additionally, NF-κB signaling suppresses apoptosis and contributes to resistance to chemotherapy and radiotherapy, thereby further emphasizing its critical role in CRC progression and therapeutic resistance [^[12]]. TNF-α is a key upstream and downstream component of inflammatory signaling in CRC development, including inflammation, tumor growth, and metastasis [^[13]]. It is a potent inducer of NF-κB activation, and its expression is further reinforced through an NF-κB-dependent positive feedback loop, thereby sustaining chronic inflammatory signaling [^[14]]. Numerous studies have demonstrated that TNF-α activates NF-κB signaling, which subsequently regulates the expression of apoptosis-related proteins, including Bcl-2 and Bax, thereby promoting tumor cell survival and inhibiting apoptosis [^[11],^[15]]. Likewise, IL-8 is an NF-κB-responsive chemokine that contributes to tumor-associated inflammation, migration, proliferation, angiogenesis, and invasive potential in colorectal cancer [^[16]]. IL-8 overexpression has been linked to resistance to chemotherapy, advanced tumor stage, lymphatic and liver metastases, and a poor prognosis [^[3]]. Therefore, suppression of TNF-α and IL-8 may reflect attenuation of NF-κB inflammatory signaling and may indicate disruption of a pro-tumorigenic cytokine in CRC [^[17]]. In addition to approved drugs, various compound scaffolds have been reported to target NF-κB, including [4,6-dichloro-N-phenyl-1,3,5-triazin-2-amine] (NI241), imidazolone derivatives, hexahydropyrido [4,3-d]pyrimidines (PPMs), and N-substituted-2-((2-oxo-2-((4-sulfamoylphenyl)amino)ethyl)thio) acetamide derivatives [^[18],^[19]]. Furthermore, the naphthoquinone derivative plumbagin (5-hydroxy-2-methyl-1,4-naphthoquinone) has been reported to suppress NF-κB activation and modulate the expression of downstream gene products by affecting p65 and IκBα kinase activity, thereby enhancing apoptosis induced by cytokines and chemotherapeutic agents [^[20]]. In oral cancer cells, shikonin (5,8-dihydroxy-2-(1-hydroxy-4-methyl-3-pentenyl)-1,4-naphthoquinone) induces apoptosis through NF-κB inhibition and subsequent caspase activation [^[21]]. A recent study reported the synthesis of anilino-1,4-naphthoquinones (Fig. 1) comprising three series of chloro-1,4-naphthoquinones (series I, 3–12), bromo-1,4-naphthoquinones (series II, 13–15), and 1,4-naphthoquinones (series III, 16–18) from 1,4-napthoquinones 1 and various aniline derivatives 2. These compounds have been reported to exhibit inhibitory activity against EGFR tyrosine kinase and tyrosinase [^[22],^[23]]. Several 1,4-naphthoquinone derivatives, including napabucasin (BBI-608), sepantronium bromide (YM-155), and menadione (vitamin K3), have been investigated for their potential therapeutic applications [^[24]]. However, the potential of anilino-1,4-naphthoquinone derivatives as NF-κB inhibitors has not yet been investigated and remains unclear.

Accordingly, NF-κB-regulated cytokine signaling may represent an important mechanistic basis for the anticancer activity of CRC. Therefore, decreased TNF-α and IL-8 expression after treatment may provide evidence that the tested compound exerts not only cytotoxic effects but also anti-inflammatory activity, which could help limit inflammation-driven colorectal carcinogenesis. The present study aimed to investigate whether the observed reduction in TNF-α and IL-8 following treatment with an anilino-1,4-naphthoquinone derivative (compound 12) is associated with an anti-inflammatory effect mediated, at least in part, through modulation of NF-κB-related signaling in colorectal cancer cells.

Materials and methods

Materials

Anilino-1,4-naphthoquinone derivatives (3-18) (Fig. 1) were previously synthesized by our research group [^[22],^[23]]. DMEM, RPMI-1640 culture medium, Fetal bovine serum (FBS), and penicillin-streptomycin were purchased from Gibco (Waltham, MA, USA). USB Corporation (Cleveland, OH, USA) provided the fluorescent dye Hoechst 33342 and substances for cell viability tests, including MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide]. For apoptosis detection, the JC-1 detection kit was purchased from Biotium (Biotium, CA, USA). Sigma-Aldrich provided the dimethyl sulfoxide (DMSO) (Merge KGaA, Darmstadt, Germany). We purchased propidium iodide (PI) from Molecular Probes (Carlsbad, CA, USA). The manufacturer of the Annexin V Apoptosis Detection Kit was BioLegend (San Diego, CA, USA). The supplier of RNase A solution was Hi-Media Laboratories (Marg, Mumbai, India).### Cell culture

Human colorectal cancer cell line, HCT116 (ATCC CCL-247) and normal Vero cells (ATCC CCL-81) were obtained from the American Type Culture Collections (ATCC, Manassas, VA, USA) and maintained according to standard culture conditions. Mycoplasma contamination was monitored by Hoechst 33342 staining as a preliminary screening assay before testing. HCT116 cells were cultured in Roswell Park Memorial Institute medium (RPMI-1640) and normal Vero cells were grown in high-glucose Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-activated FBS, 100 U/mL of penicillin and 100 μg/mL streptomycin (Gibco, Waltham, MA, USA) at 37◦C with 5% CO2 in a humidified-air incubator.### Cytotoxicity assay

The MTT assay was used to evaluate the cytotoxicity of 1,4-naphthoquinone derivatives (3–18) against HCT116 cell viability. After 24 h of cell growth, HCT116 and normal Vero cells were plated at a density of 8 × 103 cells/well on 96-well plates. The cells were co-cultured with LPS (1 μg/mL) for an additional 24 h after being pre-treated with varying concentrations of synthetic substances (0–100 μM) for 1 h. 5-FU, chemotherapeutic agent for colorectal cancer served as a reference drug control for useful benchmark biological activity. Following the incubation period, each well was filled with MTT solution (0.5 mg/mL) after the RPMI1640 and DMEM medium were removed. The cells were exposed to 5% CO2 and 37 °C for 2 h. The Synergy HT Multi-Mode microplate reader (BioTek Instruments, Inc., USA) was used to measure the absorbance (OD) at 570 nm after the formazan crystal product had been dissolved in DMSO. GraphPad software 9.0.0 (San Diego, CA, USA) was used to determine the half-maximal inhibitory concentration (IC50) values for HCT116 cells.### Nuclear morphological investigation

To determine DNA fragmentation and condensation, Hoechst 33342 staining was performed. HCT116 cells were seeded in six-well plates at a density of 1 × 105 cells per well. The selected compounds (3, 8, and 12) were added at concentrations of 12, 35, and 7 μM, respectively, and cells were pretreated for 1 h before incubation with LPS (1 μg/mL) for 48 h. 5-Fluorouracil (5-FU) was used as the reference drug. Following treatment, the cells were stained with 5 μg/mL Hoechst 33342 for 30 min at 37 °C to assess apoptotic morphological changes, including nuclear condensation and DNA fragmentation. The cells were subsequently examined and photographed using a fluorescence microscope (Olympus, Tokyo, Japan).### Mitochondrial membrane potential (ΔΨm) detection

According to the manufacturer's instructions, the JC-1 mitochondrial membrane potential detection kit (Biotium, CA, USA) was used to detect altered mitochondrial membrane potential. In summary, HCT116 cells were seeded in 12-well plates at a density of 5 × 103 cells/mL, pretreated with compounds 3, 8, and 12 at 12, 35, and 7 μM, respectively, or vehicle control (0.5% DMSO), and then co-cultured with LPS (1 μg/mL) for 24 h. The reference drug was 5-FU (25 μM). Following incubation, cells were stained with JC-1 and allowed to incubate at 37 °C in the dark for 30 min before being examined under a fluorescent microscope (DP73, Olympus, Japan).### Cell cycle detection

Propidium iodide (PI) staining (Molecular Probes, Eugene, OR, USA) was used to quantify the population of apoptotic cells in the sub-G1 phase according to the manufacturer's instructions. In brief, 6-well plates containing 3 × 105 cells/mL of HCT116 cells were cultured overnight. Cells were pretreated with compounds 3, 8, and 12 (12, 35, and 7 μM, respectively) or vehicle control (0.5% DMSO) for 1 h, followed by exposure to LPS (1 μg/mL) for 48 h. 5-Fluorouracil (5-FU; 25 μM) was used as the reference drug. After 48 h of treatment, the cells were harvested, washed with PBS, and fixed in cold 100% ethanol for at least 24 h at 4 °C. The fixed cells were then centrifuged, resuspended in RNase A (50 μg/μL in PBS), and incubated at 37 °C for 20 min. Subsequently, the cells were stained with 500 μL of propidium iodide (PI) solution for 30 min at room temperature in the dark. The sub-G1 cell population was analyzed by flow cytometry using a BD FACScan flow cytometer (Becton, Dickinson, San Jose, CA, USA). Each experiment was performed in triplicate for each treatment group.### Annexin V/PI staining for apoptosis detection

As described previously, propidium iodide (PI) staining and flow cytometry using the Annexin V Apoptosis Detection Kit with PI (BioLegend, San Diego, CA, USA) were performed to assess cell cycle distribution and quantitatively determine phosphatidylserine externalization and apoptotic cell death, respectively [^[25]]. HCT116 cells were seeded at a density of approximately 5 × 105 cells per well in 6-well plates. Following co-culture with LPS (1 μg/mL), cells were pre-treated with particular compounds 3, 8, and 12 at 12, 35, and 7 μM, respectively. 5-FU (25 μM) was then employed as a reference agent for 48 h and incubated at 37 °C with 5% CO2. Apoptotic cell death was subsequently analyzed using a FACSCanto flow cytometer (Becton Dickinson Biosciences, San Jose, CA, USA).### Molecular docking

The Human NF-κB p50 homodimer active site enzyme (PDB: 1SVC), and the p50/p65 heterodimer of transcription factor NF-κB complex (PDB: 1VKX), and human IκB kinase beta (PDB: 4KIK) were all obtained from the Protein Data Bank (PDB). The molecular docking scores of the selected best potential compound 12, predicted by the GOLD program, are compared with those of the reference drugs MLN120B and BMS-345541. Additionally, the 3D structure of the ligand compound 12 was manually constructed in BIOVIA Discovery Studio (DS), a foundational step in structure-based drug design that enables precise modeling of receptor-ligand interactions. Docking validation was performed by redocking the crystallized ligand, which was defined as the center of the active site, to evaluate ligand orientation. The Gold program with the CHEMPLP scoring function was employed for docking. The binding mode between the human NF-κB p50 homodimer active site enzyme, p50/p65 heterodimer of transcription factor NF-κB complex, human IκB kinase beta crystallized protein, tested compound 12, and NF-κB inhibitor was visualized using the BIOVIA Discovery Studio (DS) 2.5.### Molecular dynamics (MD) simulations

Two complex structures of human IκB kinase beta (IKKβ) bound to MLN120B and compound 12 were generated from the best-scoring molecular docking poses. The kinase domain of human IKKβ (residues 1–308) was selected to evaluate the binding stability of both ligands through molecular dynamics (MD) simulations. Missing residues within the kinase domain were modeled using SWISS-MODEL [^[26]]. The protonation states and side-chain charges of amino acid residues at pH 7.4 were predicted using PROPKA implemented in the APBS & PDB2PQR web server [^[27]]. The protonation states of both ligands at pH 7.4 were determined using MarvinSketch [^[28]], and the appropriate protonation forms were subsequently parameterized using the AMBER ff14SB force field [^[29]] and the Generalized AMBER force field version 2 (GAFF2) [^[30]]. Each protein–ligand complex was solvated in a truncated octahedral box of TIP3P water molecules [^[31]], extending approximately 10.0 Å from the protein surface. Counterions (Na+ or Cl−) were added to neutralize the net charge of each system. Energy minimization was performed for 40,000 steps, followed by a 2 ns equilibration phase to remove unfavorable contacts and stabilize the systems. Production MD simulations were run for 100 ns at constant pressure (1 atm) and temperature (310 K) conditions using AMBER 25 [^[32],^[33]]. The pressure was regulated using the Berendsen barostat [^[34]], while Langevin thermostat was used to keep the temperature at 310 K [^[35]]. All covalent bonds involving hydrogen atoms were constrained using the SHAKE method [^[36]]. While long-range electrostatic interactions were handled using the particle mesh Ewald (PME) method using rapid Fourier transform algorithms, van der Waals interactions were computed using a cutoff distance of 10 Å [^[37]].

The time-dependent profiles of (i) the root mean square deviation (RMSD) of the ligand and the backbone atoms of protein residues within 5 Å of the ligand; (ii) the solvent-accessible surface area (SASA) of amino acid residues within 5 Å of the ligand in the ATP-binding pocket; (iii) the total number of atomic contacts (#contacts), defined as all heavy-atom interactions within 5 Å between the kinase domain and the ligand; and (iv) the total number of hydrogen bonds (#H-bonds) in the protein–ligand complexes were analyzed to assess the stability of protein–ligand interactions. H-bond was identified using a donor–acceptor heavy atom distance cutoff of 3.5 Å and a donor–hydrogen–acceptor (D–H···A) angle ranging from 120° to 180°. All analyses were performed using the CPPTRAJ module [^[38]] implemented in the AMBER software suite. The binding affinities between the kinase domain and each ligand were estimated from MD simulation snapshots collected during the final 20 ns (80–100 ns) using the MM/GBSA approach [^[39],^[40]]. The overall binding free energy () and per-residue decomposition free energy () were calculated using the MMPBSA. py [^[41]]. Structural visualization and graphical representation of protein–ligand complexes obtained from the MD simulations were generated using UCSF ChimeraX [^[42]].### Cytokine secretion measurement

Cytokine detection was determined according to the manufacturer's instructions. Briefly, HCT116 cells (3 × 105 cells/well) were pre-treated for 1 h with selected compound 12 at 3.5, 7, and 14 μM, or reference drug rofecoxib at the indicated concentrations, followed by co-culture with 1 μg/mL of LPS for 48 h. TNF-α and IL-8 levels in culture supernatants were measured using multiplex xMAP technology with the MILLIPLEX Human Cytokine/Chemokine/Growth Factor Panel A immunoassay ELISA kit (HCYTA-60 K, Merck, Darmstadt, Germany). The multiplex assay was conducted on the MAGPIX system (Merck, Darmstadt, Germany), and data analysis was performed using MILLIPLEX Analyst v5.1 software (Merck). Each sample was measured in duplicate.### Statistical analysis

Statistical analysis was performed using GraphPad Prism version 9.0.0 (GraphPad Software, San Diego, CA, USA). One-way analysis of variance (ANOVA) followed by Dunnett's multiple comparison test. Data are presented as the mean ± SD. A p-value of *p˂0.05, p˂0.01, *p < 0.001, and ****p < 0.001 were considered statistically significant (n = 3).

Materials

Anilino-1,4-naphthoquinone derivatives (3-18) (Fig. 1) were previously synthesized by our research group [^[22],^[23]]. DMEM, RPMI-1640 culture medium, Fetal bovine serum (FBS), and penicillin-streptomycin were purchased from Gibco (Waltham, MA, USA). USB Corporation (Cleveland, OH, USA) provided the fluorescent dye Hoechst 33342 and substances for cell viability tests, including MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide]. For apoptosis detection, the JC-1 detection kit was purchased from Biotium (Biotium, CA, USA). Sigma-Aldrich provided the dimethyl sulfoxide (DMSO) (Merge KGaA, Darmstadt, Germany). We purchased propidium iodide (PI) from Molecular Probes (Carlsbad, CA, USA). The manufacturer of the Annexin V Apoptosis Detection Kit was BioLegend (San Diego, CA, USA). The supplier of RNase A solution was Hi-Media Laboratories (Marg, Mumbai, India).

Cell culture

Human colorectal cancer cell line, HCT116 (ATCC CCL-247) and normal Vero cells (ATCC CCL-81) were obtained from the American Type Culture Collections (ATCC, Manassas, VA, USA) and maintained according to standard culture conditions. Mycoplasma contamination was monitored by Hoechst 33342 staining as a preliminary screening assay before testing. HCT116 cells were cultured in Roswell Park Memorial Institute medium (RPMI-1640) and normal Vero cells were grown in high-glucose Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% heat-activated FBS, 100 U/mL of penicillin and 100 μg/mL streptomycin (Gibco, Waltham, MA, USA) at 37◦C with 5% CO2 in a humidified-air incubator.

Cytotoxicity assay

The MTT assay was used to evaluate the cytotoxicity of 1,4-naphthoquinone derivatives (3–18) against HCT116 cell viability. After 24 h of cell growth, HCT116 and normal Vero cells were plated at a density of 8 × 103 cells/well on 96-well plates. The cells were co-cultured with LPS (1 μg/mL) for an additional 24 h after being pre-treated with varying concentrations of synthetic substances (0–100 μM) for 1 h. 5-FU, chemotherapeutic agent for colorectal cancer served as a reference drug control for useful benchmark biological activity. Following the incubation period, each well was filled with MTT solution (0.5 mg/mL) after the RPMI1640 and DMEM medium were removed. The cells were exposed to 5% CO2 and 37 °C for 2 h. The Synergy HT Multi-Mode microplate reader (BioTek Instruments, Inc., USA) was used to measure the absorbance (OD) at 570 nm after the formazan crystal product had been dissolved in DMSO. GraphPad software 9.0.0 (San Diego, CA, USA) was used to determine the half-maximal inhibitory concentration (IC50) values for HCT116 cells.

Nuclear morphological investigation

To determine DNA fragmentation and condensation, Hoechst 33342 staining was performed. HCT116 cells were seeded in six-well plates at a density of 1 × 105 cells per well. The selected compounds (3, 8, and 12) were added at concentrations of 12, 35, and 7 μM, respectively, and cells were pretreated for 1 h before incubation with LPS (1 μg/mL) for 48 h. 5-Fluorouracil (5-FU) was used as the reference drug. Following treatment, the cells were stained with 5 μg/mL Hoechst 33342 for 30 min at 37 °C to assess apoptotic morphological changes, including nuclear condensation and DNA fragmentation. The cells were subsequently examined and photographed using a fluorescence microscope (Olympus, Tokyo, Japan).

Mitochondrial membrane potential (ΔΨm) detection

According to the manufacturer's instructions, the JC-1 mitochondrial membrane potential detection kit (Biotium, CA, USA) was used to detect altered mitochondrial membrane potential. In summary, HCT116 cells were seeded in 12-well plates at a density of 5 × 103 cells/mL, pretreated with compounds 3, 8, and 12 at 12, 35, and 7 μM, respectively, or vehicle control (0.5% DMSO), and then co-cultured with LPS (1 μg/mL) for 24 h. The reference drug was 5-FU (25 μM). Following incubation, cells were stained with JC-1 and allowed to incubate at 37 °C in the dark for 30 min before being examined under a fluorescent microscope (DP73, Olympus, Japan).

Cell cycle detection

Propidium iodide (PI) staining (Molecular Probes, Eugene, OR, USA) was used to quantify the population of apoptotic cells in the sub-G1 phase according to the manufacturer's instructions. In brief, 6-well plates containing 3 × 105 cells/mL of HCT116 cells were cultured overnight. Cells were pretreated with compounds 3, 8, and 12 (12, 35, and 7 μM, respectively) or vehicle control (0.5% DMSO) for 1 h, followed by exposure to LPS (1 μg/mL) for 48 h. 5-Fluorouracil (5-FU; 25 μM) was used as the reference drug. After 48 h of treatment, the cells were harvested, washed with PBS, and fixed in cold 100% ethanol for at least 24 h at 4 °C. The fixed cells were then centrifuged, resuspended in RNase A (50 μg/μL in PBS), and incubated at 37 °C for 20 min. Subsequently, the cells were stained with 500 μL of propidium iodide (PI) solution for 30 min at room temperature in the dark. The sub-G1 cell population was analyzed by flow cytometry using a BD FACScan flow cytometer (Becton, Dickinson, San Jose, CA, USA). Each experiment was performed in triplicate for each treatment group.

Annexin V/PI staining for apoptosis detection

As described previously, propidium iodide (PI) staining and flow cytometry using the Annexin V Apoptosis Detection Kit with PI (BioLegend, San Diego, CA, USA) were performed to assess cell cycle distribution and quantitatively determine phosphatidylserine externalization and apoptotic cell death, respectively [^[25]]. HCT116 cells were seeded at a density of approximately 5 × 105 cells per well in 6-well plates. Following co-culture with LPS (1 μg/mL), cells were pre-treated with particular compounds 3, 8, and 12 at 12, 35, and 7 μM, respectively. 5-FU (25 μM) was then employed as a reference agent for 48 h and incubated at 37 °C with 5% CO2. Apoptotic cell death was subsequently analyzed using a FACSCanto flow cytometer (Becton Dickinson Biosciences, San Jose, CA, USA).

Molecular docking

The Human NF-κB p50 homodimer active site enzyme (PDB: 1SVC), and the p50/p65 heterodimer of transcription factor NF-κB complex (PDB: 1VKX), and human IκB kinase beta (PDB: 4KIK) were all obtained from the Protein Data Bank (PDB). The molecular docking scores of the selected best potential compound 12, predicted by the GOLD program, are compared with those of the reference drugs MLN120B and BMS-345541. Additionally, the 3D structure of the ligand compound 12 was manually constructed in BIOVIA Discovery Studio (DS), a foundational step in structure-based drug design that enables precise modeling of receptor-ligand interactions. Docking validation was performed by redocking the crystallized ligand, which was defined as the center of the active site, to evaluate ligand orientation. The Gold program with the CHEMPLP scoring function was employed for docking. The binding mode between the human NF-κB p50 homodimer active site enzyme, p50/p65 heterodimer of transcription factor NF-κB complex, human IκB kinase beta crystallized protein, tested compound 12, and NF-κB inhibitor was visualized using the BIOVIA Discovery Studio (DS) 2.5.

Molecular dynamics (MD) simulations

Two complex structures of human IκB kinase beta (IKKβ) bound to MLN120B and compound 12 were generated from the best-scoring molecular docking poses. The kinase domain of human IKKβ (residues 1–308) was selected to evaluate the binding stability of both ligands through molecular dynamics (MD) simulations. Missing residues within the kinase domain were modeled using SWISS-MODEL [^[26]]. The protonation states and side-chain charges of amino acid residues at pH 7.4 were predicted using PROPKA implemented in the APBS & PDB2PQR web server [^[27]]. The protonation states of both ligands at pH 7.4 were determined using MarvinSketch [^[28]], and the appropriate protonation forms were subsequently parameterized using the AMBER ff14SB force field [^[29]] and the Generalized AMBER force field version 2 (GAFF2) [^[30]]. Each protein–ligand complex was solvated in a truncated octahedral box of TIP3P water molecules [^[31]], extending approximately 10.0 Å from the protein surface. Counterions (Na+ or Cl−) were added to neutralize the net charge of each system. Energy minimization was performed for 40,000 steps, followed by a 2 ns equilibration phase to remove unfavorable contacts and stabilize the systems. Production MD simulations were run for 100 ns at constant pressure (1 atm) and temperature (310 K) conditions using AMBER 25 [^[32],^[33]]. The pressure was regulated using the Berendsen barostat [^[34]], while Langevin thermostat was used to keep the temperature at 310 K [^[35]]. All covalent bonds involving hydrogen atoms were constrained using the SHAKE method [^[36]]. While long-range electrostatic interactions were handled using the particle mesh Ewald (PME) method using rapid Fourier transform algorithms, van der Waals interactions were computed using a cutoff distance of 10 Å [^[37]].

The time-dependent profiles of (i) the root mean square deviation (RMSD) of the ligand and the backbone atoms of protein residues within 5 Å of the ligand; (ii) the solvent-accessible surface area (SASA) of amino acid residues within 5 Å of the ligand in the ATP-binding pocket; (iii) the total number of atomic contacts (#contacts), defined as all heavy-atom interactions within 5 Å between the kinase domain and the ligand; and (iv) the total number of hydrogen bonds (#H-bonds) in the protein–ligand complexes were analyzed to assess the stability of protein–ligand interactions. H-bond was identified using a donor–acceptor heavy atom distance cutoff of 3.5 Å and a donor–hydrogen–acceptor (D–H···A) angle ranging from 120° to 180°. All analyses were performed using the CPPTRAJ module [^[38]] implemented in the AMBER software suite. The binding affinities between the kinase domain and each ligand were estimated from MD simulation snapshots collected during the final 20 ns (80–100 ns) using the MM/GBSA approach [^[39],^[40]]. The overall binding free energy () and per-residue decomposition free energy () were calculated using the MMPBSA. py [^[41]]. Structural visualization and graphical representation of protein–ligand complexes obtained from the MD simulations were generated using UCSF ChimeraX [^[42]].

Cytokine secretion measurement

Cytokine detection was determined according to the manufacturer's instructions. Briefly, HCT116 cells (3 × 105 cells/well) were pre-treated for 1 h with selected compound 12 at 3.5, 7, and 14 μM, or reference drug rofecoxib at the indicated concentrations, followed by co-culture with 1 μg/mL of LPS for 48 h. TNF-α and IL-8 levels in culture supernatants were measured using multiplex xMAP technology with the MILLIPLEX Human Cytokine/Chemokine/Growth Factor Panel A immunoassay ELISA kit (HCYTA-60 K, Merck, Darmstadt, Germany). The multiplex assay was conducted on the MAGPIX system (Merck, Darmstadt, Germany), and data analysis was performed using MILLIPLEX Analyst v5.1 software (Merck). Each sample was measured in duplicate.

Statistical analysis

Statistical analysis was performed using GraphPad Prism version 9.0.0 (GraphPad Software, San Diego, CA, USA). One-way analysis of variance (ANOVA) followed by Dunnett's multiple comparison test. Data are presented as the mean ± SD. A p-value of *p˂0.05, p˂0.01, *p < 0.001, and ****p < 0.001 were considered statistically significant (n = 3).

Results and discussion

Biological processes

Antiproliferative effects of 1,4-naphthoquinone derivatives (3-18) on HCT116 cells

The cellular toxicity effects of 1,4-naphthoquinone derivatives (3-18) and the reference drug, 5-fluorouracil (5-FU) on HCT116 colorectal cancer cells and the normal African green monkey (Vero cells) cell line were determined using the MTT assay. As a result, 5-FU exhibited relatively weak cytotoxicity against HCT116 cells, while showing greater toxicity toward normal Vero cells than the corresponding 1,4-naphthoquinone derivatives. Previous studies have reported that 1,4-naphthoquinone derivatives possess anticancer activity against various cancer types, including lung, liver, breast, and colorectal cancers, while exhibiting relatively low cytotoxicity toward normal human embryonic lung MRC-5 cells [^[22]]. These findings are consistent with our experimental results. HCT116 cells were treated with the synthesized compounds (3–18) at concentrations ranging from 0 to 100 μM for 48 h, as indicated in Table 1. Among the tested compound, 2-chloro-3-(p-tolylamino)naphthalene-1,4-dione 3, 2-chloro-3-((4-nitrophenyl)amino)naphthalene-1,4-dione 8 and 2-chloro-3-((4-hydroxyphenyl)amino)naphthalene-1,4-dione 12 (Fig. 2) exhibited the strongest anticancer activity, as evidenced by their dose-dependent inhibition of HCT116 cell proliferation. The antiproliferative activity followed the order 12 > 3 > 8, corresponding to the substituent effects of R = 4-OH > 4-CH3 > 4-NO2). The IC50 value of compounds 12, 3, and 8 were 7.32 ± 0.45, 12.57 ± 2.66, and 36.54 ± 1.02 μM, respectively (Table 1). The superior activity of the compound 12 (R = 4-OH) compared with derivatives 3 (R = 4-CH3) and 8 (R = 4-NO2) can be attributed to its ability to form hydrogen-bonding interaction with amino acid residue within the target binding site. Furthermore, the hydroxyl group provides electron-donating character through resonance, which can optimize the electronic distribution of the aromatic ring and favor interactions with the receptor. The study suggests that the hydroxyl substituent provides an optimal balance of electronic modulation and hydrogen-bonding interaction, resulting in superior potency and improved selectivity. Remarkably, compound 12 emerged as the most promising candidate, exhibiting potent antiproliferative activity and greater selectivity toward HCT116 colorectal cancer cells than 5-FU, which was used as the positive pharmacological control. Compound 12 showed an IC50 value of 7.32 ± 0.45 μM against HCT116 cells, whereas 5-FU exhibited an IC50 value of 52.58 ± 1.24 μM. In contrast, most of the synthesized compounds showed relatively low cytotoxicity toward normal Vero cells, while 5-FU exhibited substantial cytotoxicity, with an IC50 value of 6.07 ± 1.22 μM. Notably, compound 12 exhibited a selectivity index (SI) of approximately 7, which was approximately 57-fold higher than that of 5-FU (SI = 0.12), indicating a markedly greater selectivity for cancer cells over normal cells. Therefore, although compound 12 demonstrated potent inhibitory activity against HCT116 cell growth, further mechanistic studies are warranted to elucidate its underlying molecular mechanisms.### Effect of 1,4-naphthoquinone derivatives (compounds 3, 8, and 12) on apoptotic morphology and mitochondrial membrane potential (ΔΨm)

To investigate whether the antiproliferative effects of compounds 3, 8, and 12 were associated with apoptotic characteristics, including nuclear morphological alterations, DNA fragmentation, and DNA pyknosis. Hoechst 33342 staining was performed to characterize these features in treated HCT116 cells. Hoechst 33342 staining is a widely used method for evaluating apoptotic morphological changes because it enables the distinction between normal and apoptotic nuclei based on characteristic alterations in nuclear morphology and chromatin structure following exposure to drugs or test compounds. These apoptotic features include chromatin condensation, nuclear shrinkage, nuclear fragmentation, and the formation of apoptotic bodies, which can be visualized by fluorescence microscopy [^[43]]. As shown in Fig. 3, the morphological changes in the HCT116 cells that were pre-exposed to compounds 3, 8, and 12 (12, 35, and 7 μM, respectively) for 1 h, followed by co-culture with 1 μg/mL LPS for 48 h, exhibited significant morphological changes characteristic of apoptosis. Following 48 h, they included chromatin condensation, cell shrinkage, DNA fragmentation, and the formation of apoptotic bodies, which were comparable to those seen with the reference drug, 5-FU (25 μM). For Fig. 4, JC-1 staining revealed that compounds 3, 8, and 12 induced a loss of mitochondrial membrane potential (ΔΨm) in HCT116 cells, suggesting mitochondrial dysfunction and activation of the intrinsic apoptotic pathway. In untreated or LPS-treated cells with high ΔΨm, JC-1 accumulated within mitochondria to form J-aggregates that emitted red fluorescence, whereas cells with a reduced ΔΨm exhibited predominantly green fluorescence [^[44]]. Our results demonstrated that treatment with compounds 3, 8, and 12 for 24 h significantly reduced ΔΨm, as indicated by decreased red fluorescence and increased green fluorescence. Among the tested compounds, compound 12 produced the most pronounced increase in green fluorescence compared with the LPS-treated group and the reference drug 5-FU, indicating a greater disruption of mitochondrial function. Collectively, these findings suggest that the selected compounds differentially affect mitochondrial stability, with compound 12 exhibiting the strongest effect on mitochondrial membrane potential. The loss of ΔΨm may contribute to the induction of apoptosis and the antiproliferative effects observed in HCT116 cells. Compounds 3 and 8 also induced mitochondrial depolarization, although their effects appeared to be less pronounced than those of compound 12. These findings support the involvement of mitochondrial dysfunction in the apoptotic effects of compounds 3, 8, and 12 in HCT116 cells.### Effect of 1,4-naphthoquinone derivatives (compounds 3, 8, and 12) treatment on cell cycle distribution

In order to assess the significance of compounds 3, 8, and 12 at 12, 35, and 7 μM, respectively, on the induction of apoptosis, HCT116 cells were stained with propidium iodide (PI), and flow cytometry was used to measure DNA content and cell cycle distribution. After 1 h of pretreatment with these compounds or the reference drug, 5-FU, HCT116 cells were co-cultured with 1 μg/mL LPS for 48 h (Fig. 5). After 48 h of treatment, HCT116 cells were exposed to compound 3 (12 μM), compound 8 (35 μM), compound 12 (7 μM), or 5-FU (25 μM), corresponding to their respective IC50 concentrations. As shown in Fig. 5, the relative distribution of HCT116 cells across the different cell cycle phases revealed an increase in the sub-G1 population following treatment with compounds 3, 8, and 12. Among the tested compounds, compound 12 produced the most pronounced increase in the sub-G1 population, from 1.50 ± 0.61% in untreated control cells and 1.60 ± 0.10% in LPS-treated cells to 19.03 ± 5.25% following treatment with compound 12. In comparison, 5-FU increased the sub-G1 population to 17.40 ± 3.90% (Fig. 5A and B and Table 2). These findings suggest that compound 12 induces DNA fragmentation and apoptotic cell death in HCT116 cells, as reflected by the marked accumulation of cells in the sub-G1 population. Notably, compound 12 exhibited a stronger effect on sub-G1 accumulation than 5-FU after 48 h of treatment, suggesting a greater potential to induce apoptotic cell death under the conditions tested.### Effect of 1,4-naphthoquinone derivatives (compounds 3, 8, and 12) treatment on apoptosis

As shown in Fig. 6, flow cytometric analysis revealed that pretreatment of HCT116 cells with compounds 3, 8, and 12 at concentrations of 12, 35, and 7 μM, respectively, for 1 h, followed by exposure to LPS (1 μg/mL) for 48 h, resulted in varying degrees of apoptotic cell death. The untreated control group exhibited a low proportion of apoptotic cells, indicating a minimal basal level of apoptosis, which was comparable to that observed in the LPS-treated group. Although LPS treatment resulted in a slight increase in the apoptotic cell population, the difference was not statistically significant. Treatment with compounds 3 and 8 in the presence of LPS resulted in a moderate increase in the apoptotic cell population (Fig. 6A and B).

In contrast, the LPS/compound 12 (7 μM) treatment produced the most pronounced induction of apoptosis in HCT116 cells. Compound 12 increased the early apoptotic population to 8.40 ± 0.82% and the late apoptotic population to 46.33 ± 5.71%, compared with 8.27 ± 1.46% for the late apoptotic population in the LPS-treated control group (1 μg/mL). The marked increase in the late apoptotic population following compound 12 treatment indicates a substantial induction of apoptotic cell death. As shown in Fig. 6C, the heatmap provides a visual representation of these changes, with the strongest signal corresponding to the late apoptotic population following treatment with compound 12, whereas the increase in early apoptosis was comparatively modest. Treatment with 5-FU (25 μM), used as the reference drug, also increased apoptosis in HCT116 cells; however, the late apoptotic population was lower than that observed following treatment with compound 12. Collectively, these findings demonstrate that compound 12 effectively induces apoptotic cell death in HCT116 cells.### Binding interactions predicted by molecular docking

Molecular docking was conducted to investigate the binding interactions of the chosen compound 12, which exhibited the highest potent biological activity against human HCT116 cells. To prove the anti-inflammatory properties of compound 12, a computational docking simulation (in silico docking) of its interactions with NF-κB was performed. As shown in Fig. 7, Fig. 8, Fig. 9. Human NF-κB p50 homodimer (PDB ID: 1SVC), p50/p65 heterodimer NF-κB complex (PDB ID: 1VKX), and human IκB kinase β (PDB ID: 4KIK). The procedure for docking was successfully validated by docking the co-crystallized ligand into the NF-κB binding site using the GOLD program with the CHEMPLP scoring function. The results of molecular docking analysis demonstrated that this compound showed potential as an NF-κB inhibitor, with fitness scores similar to those of the reference drugs MLN120B and BMS-345541 (Table 3). Moreover, compound 12 could also be favorably accommodated within functionally relevant regions of the NF-κB signaling axis, including human NF-κB/p50 (Fig. 7), the p50/p65 NF-κB complex (Fig. 8), and human Ikβ kinase (Fig. 9), with binding patterns that were broadly comparable to those of the reference inhibitors MLN120B, a potent active inhibitor of IKKβ, and BMS345541, a highly selective IKK-1/IKK-2 inhibitor. In the human NF-κB/p50 model, the results showed that compound 12 was positioned near the N-terminal DNA-binding region and linker-associated surface, where it formed an interaction involving a hydrogen bond with Met208, Ser211, and Tyr60 and hydrophobic contacts with residues such as Lys147, Leu210, Lys244, Ala245, and Pro246, suggesting stable occupancy of a pocket associated with DNA recognition and transcriptional regulation. These simulations are mechanistically relevant because the p50 subunit contains two domains connected by a flexible linker, and residues in this area have been implicated in ligand recognition and modulation of NF-κB/DNA interactions [^[45]]. As shown in Fig. 8. In the p50/p65 NF-κB complex, compound 12 also occupied a binding region adjacent to the DNA-interacting interface. It established several residues interaction in active site with hydrogen bond, Halogen, van der Waals, and alkyl/π-alkyl contacts with surrounding residues, including Gly365, Val358, Gly438, His364, Ser363, Pro362, Arg356, Gly361, Val412, and Leu440, indicating that the compound may interfere with heterodimer-associated DNA binding or local conformational stabilization due to DNA binding and dimer integrity are critical for NF-κB transcriptional activity. Therefore, the interaction effect of compound 12 in this NF-κB active site, which leading to inhibitory effect on downstream inflammatory gene expression. The similarity of the binding pose of compound 12 to MLN120B and BMS345541 in this modeling study further indicates that compound 12 can bind to pharmacologically important surfaces within the NF-κB complex and may have a further inhibitory effect on inflammation [^[46]]. As shown in Fig. 9, docking against human IKβ kinase β revealed that compound 12 could also bind within the kinase-associated pocket of the upstream regulatory enzyme, with interactions involving residues such as Cys99, Gly102, Ala42, Val29, Ile165, Met96, Leu21, Glu97, Val74, and Tyr98. These contacts included conventional hydrogen bonds, carbon-hydrogen interactions, halogen interactions, and multiple hydrophobic contacts, indicating favorable compatibility between the ligand scaffold and the IKKβ binding active site. This finding importantly suggests that IKKβ is a principal upstream activator of canonical NF-κB signaling by phosphorylating IκB, leading to its degradation. Surprisingly, previous studies showed that the reference drug, MLN120B targets IKKβ to inhibit TNF-α-induced NF-κB activation [^[45]]. Therefore, the docking profile of selected compound 12 suggests that it could influence the NF-κB pathway at multiple cellular levels, potentially affecting both upstream kinase regulation and downstream transcription factor function.### Molecular dynamics study of human IκB kinase beta (IKKβ) in complex with compound 12

The binding stability of compound 12 within the kinase domain of human IKKβ was further investigated using MD simulations, with MLN120B employed as a reference inhibitor. The RMSD time profiles demonstrated lower structural deviations for compound 12 compared with MLN120B in both the ligand and the protein backbone residues located within 5 Å of the ligand (Fig. 10A). During the equilibrated 80-100 ns interval of the MD trajectories, the average RMSD values of the protein backbone were 1.87 ± 0.19 Å for the MLN120B complex and 0.64 ± 0.08 Å for the 2-chloro-3-((4-hydroxyphenyl)amino)naphthalene-1,4-dione complex, as well as the corresponding ligand RMSD values were 1.44 ± 0.47 Å and 0.48 ± 0.12 Å, respectively. These findings suggest that the complex containing compound 12 exhibited greater conformational stability, consistent with SASA and H-bond analyses. The higher structural fluctuation observed in the MLN120B complex was associated with greater solvent exposure around the binding pocket, as reflected by its higher SASA value (705 ± 75 Å2) relative to compound 12 (529 ± 48 Å2). The enhanced stability of compound 12 may be attributed to its higher #H-bonds compared with MLN120B, which was related to differences in ligand-binding orientation (Fig. 10B). In contrast, MLN120B exhibited a greater number of atomic contacts (#contacts = 235 ± 23) than compound 12 (#contacts = 209 ± 17), likely due to its larger molecular size, which provided a broader surface area for intermolecular interactions. Overall, these results suggest that compound 12 forms a more stable binding complex within the kinase domain of human IKKβ than MLN120B.

The MM/GBSA method was used to further assess the binding affinities of the protein–ligand complexes. Table 4 shows that both compounds exhibited comparable binding free energies (), with values of −29.89 ± 0.23 kcal/mol for MLN120B and −29.95 ± 0.14 kcal/mol for compound 12. The major contribution to the vacuum binding free energy () arose from van der Waals interactions (), whereas electrostatic interactions () contributed to a lesser extent, as also observed by other inhibitors in a previous study [^[47]]. This behavior is consistent with the predominance of hydrophobic hotspot residues within the binding pocket (Fig. 9) and the largely neutral characteristics of both inhibitors. Per-residue free energy decomposition () analysis identified key binding residues with stabilizing interaction energies lower than −1.0 kcal/mol (Fig. 10C, top). Both ligands shared several common hotspot residues, including Leu21, Val29, Gly102, Val152, and Ile165. Notably, compound 12 additionally exhibited significant interactions with Tyr98 and Cys99, suggesting that despite differences in chemical structure and binding geometry, compound 12 retained a binding mode similar to that of the reference inhibitor while also establishing additional stabilizing interactions with neighboring residues (Fig. 10C, bottom). The strong energy contribution of Cys99 (−3.45 kcal/mol) in the compound 12 complex can be attributed to the generation of two highly persistent H-bonds between the backbone atoms of Cys99 and the carbonyl group (98% occupancy) and −NH− group (100% occupancy) of compound 12 (Fig. 10D). These findings are consistent with the RMSD, SASA, and #H-bonds, which collectively suggested enhanced stability of the compound 12 complex. In contrast, MLN120B formed only a relatively weak H-bond with Lys106, which was not identified as a hotspot residue. Nevertheless, the larger molecular size of MLN120B and its greater #contacts enabled the complex to maintain an overall binding affinity comparable to that of compound 12.

The ATP-binding pocket located in the kinase domain of human IKKβ has been classified into several functional regions [^[48]]. The first region is the glycine-rich loop (G-loop or P-loop), which includes Leu21 and Val29. The second region corresponds to the hinge region, comprising Tyr98, Cys99, and Gly102. Val152 is located on the β-strand preceding the catalytic loop, whereas Ile165 is positioned at the beginning of the catalytic loop. Previous mutagenesis studies investigating the role of Cys99 demonstrated that the C99S mutation reduced the binding affinity of MMPP, an ATP-competitive inhibitor, toward human IKKβ [^[49]], highlighting how important this residue is for ligand recognition and stabilization. Notably, the majority of the critical binding residues found in our investigation for both MLN120B and 2-chloro-3-((4-hydroxyphenyl)amino)naphthalene-1,4-dione have also been reported in previous investigations of potential human IKKβ inhibitors, including 4-phenyl-7-azaindoles [^[50]], rotigotine [^[51]], 4-phenyl-5-p-tolyloxymethyl-4H-1,2,4-triazole thioethers [^[52]], and coumestrol [^[53]]. These observations support the reliability and consistency of the present in silico analyses with previously reported structure–activity relationships of IKKβ inhibitors. Collectively, these findings highlight compound 12 as promising inhibitors of human IKKβ.### Effect of 1,4-naphthoquinone derivatives (compounds3,8, and12) treatment inhibits TNF-α and IL-8 inflammatory cytokine release

Molecular docking analysis of compound 12 revealed favorable binding interactions with specific NF-κB target sites, involving key amino acid residues that were also engaged by the reference inhibitors MLN120B and BMS-345541. These findings suggest that compound 12 may interfere with the structural and functional activity of NF-κB-related target proteins. To further investigate the biological relevance of NF-κB-associated signaling, the secretion of inflammatory cytokines was subsequently quantified using the MILLIPLEX assay. As shown in Fig. 11, the inflammatory cytokines evaluated in this study were TNF-α (Fig. 11A) and IL-8 (Fig. 11B), both of which are widely investigated as mediators of inflammatory responses and potential targets for anti-inflammatory intervention [^[54]]. TNF-α is a key activator of the canonical NF-κB pathway and participates in a positive feedback loop that can sustain NF-κB-mediated inflammatory signaling. Upon binding to its receptor, TNF-α activates the IKK complex, leading to IκB degradation and subsequent nuclear translocation of NF-κB dimers, including the RelA/p50 complex, where they promote the transcription of genes involved in inflammation and cell survival. Therefore, the statistically significant decrease in TNF-α production following treatment with compound 12 may reflect attenuation of the inflammatory feedback mechanisms that contribute to sustained NF-κB activation in colon cancer cells [^[5]]. IL-8, an NF-κB-regulated chemokine, also plays an essential role in maintaining a pro-inflammatory tumor microenvironment by promoting immune cell recruitment, angiogenesis, and cancer cell invasion. Previous studies have reported elevated IL-8 expression in colorectal cancer tissues compared with normal mucosa and have implicated IL-8 in interactions between tumor cells and stromal and immune cells within the tumor microenvironment. Therefore, the decrease in IL-8 after treatment with compound 12 is more significant than a simple reduction in a general inflammatory marker and may reflect the capacity of these compounds to lessen the possibility of a microenvironment conducive to cancer proliferation [^[55]]. Interestingly, the results demonstrated that the synthesized chemical dramatically decreased TNF-α and IL-8 release in HCT116 cells following LPS treatment in a dose-dependent manner, compared to the control group and standard drugs. These results suggest that compound 12 may exert anti-inflammatory and anticancer effects through suppression of the NF-κB-associated cytokine network. Importantly, these findings suggest that the synthetic compound not only binds to specific enzyme sites but also exhibits biological effects consistent with the attenuation of NF-κB-associated inflammatory signaling, which may contribute to its potential anticancer activity against colorectal cancer. Although the observed decrease in TNF-α and IL-8 is consistent with suppression of inflammatory signaling, direct biochemical evidence for NF-κB or IKKβ inhibition was not obtained in the present study. Therefore, the proposed mechanism should be regarded as preliminary and requires further validation.

Additionally, compound 12 showed significant anticancer activity in our investigation, including growth suppression and apoptosis initiation in colorectal cancer cells. Moreover, our independent research in HCT116 cells demonstrated that the identical compound suppressed the expression of the inflammatory mediators TNF-α and IL-8, while also promoting apoptotic cell death, confirming our hypothesis that compound 12 could function via both anti-inflammatory and pro-apoptotic pathways. Naphthoquinone derivatives are acknowledged as redox-active molecules that can produce reactive oxygen species (ROS), a characteristic often linked to their anticancer efficacy via the promotion of oxidative stress, mitochondrial impairment, and apoptosis in cancerous cells [[^[56]], [^[57]], [^[58]]]. However, the same redox reactivity may also raise concerns regarding off-target toxicity, particularly in non-malignant cells or tissues with limited antioxidant capacity [^[59],^[60]]. Therefore, although the current results confirm the anticancer efficacy of the produced naphthoquinone derivatives, their therapeutic use should be approached with caution and not presumed to guarantee universal safety. However, due to the potential for ROS-induced cytotoxicity to create off-target effects in non-cancerous cells, additional research is necessary to elucidate precisely the mechanism of action of compound 12 and to evaluate its selectivity and safety in relevant normal-cell and in vivo models.

Biological processes

Antiproliferative effects of 1,4-naphthoquinone derivatives (3-18) on HCT116 cells

The cellular toxicity effects of 1,4-naphthoquinone derivatives (3-18) and the reference drug, 5-fluorouracil (5-FU) on HCT116 colorectal cancer cells and the normal African green monkey (Vero cells) cell line were determined using the MTT assay. As a result, 5-FU exhibited relatively weak cytotoxicity against HCT116 cells, while showing greater toxicity toward normal Vero cells than the corresponding 1,4-naphthoquinone derivatives. Previous studies have reported that 1,4-naphthoquinone derivatives possess anticancer activity against various cancer types, including lung, liver, breast, and colorectal cancers, while exhibiting relatively low cytotoxicity toward normal human embryonic lung MRC-5 cells [^[22]]. These findings are consistent with our experimental results. HCT116 cells were treated with the synthesized compounds (3–18) at concentrations ranging from 0 to 100 μM for 48 h, as indicated in Table 1. Among the tested compound, 2-chloro-3-(p-tolylamino)naphthalene-1,4-dione 3, 2-chloro-3-((4-nitrophenyl)amino)naphthalene-1,4-dione 8 and 2-chloro-3-((4-hydroxyphenyl)amino)naphthalene-1,4-dione 12 (Fig. 2) exhibited the strongest anticancer activity, as evidenced by their dose-dependent inhibition of HCT116 cell proliferation. The antiproliferative activity followed the order 12 > 3 > 8, corresponding to the substituent effects of R = 4-OH > 4-CH3 > 4-NO2). The IC50 value of compounds 12, 3, and 8 were 7.32 ± 0.45, 12.57 ± 2.66, and 36.54 ± 1.02 μM, respectively (Table 1). The superior activity of the compound 12 (R = 4-OH) compared with derivatives 3 (R = 4-CH3) and 8 (R = 4-NO2) can be attributed to its ability to form hydrogen-bonding interaction with amino acid residue within the target binding site. Furthermore, the hydroxyl group provides electron-donating character through resonance, which can optimize the electronic distribution of the aromatic ring and favor interactions with the receptor. The study suggests that the hydroxyl substituent provides an optimal balance of electronic modulation and hydrogen-bonding interaction, resulting in superior potency and improved selectivity. Remarkably, compound 12 emerged as the most promising candidate, exhibiting potent antiproliferative activity and greater selectivity toward HCT116 colorectal cancer cells than 5-FU, which was used as the positive pharmacological control. Compound 12 showed an IC50 value of 7.32 ± 0.45 μM against HCT116 cells, whereas 5-FU exhibited an IC50 value of 52.58 ± 1.24 μM. In contrast, most of the synthesized compounds showed relatively low cytotoxicity toward normal Vero cells, while 5-FU exhibited substantial cytotoxicity, with an IC50 value of 6.07 ± 1.22 μM. Notably, compound 12 exhibited a selectivity index (SI) of approximately 7, which was approximately 57-fold higher than that of 5-FU (SI = 0.12), indicating a markedly greater selectivity for cancer cells over normal cells. Therefore, although compound 12 demonstrated potent inhibitory activity against HCT116 cell growth, further mechanistic studies are warranted to elucidate its underlying molecular mechanisms.### Effect of 1,4-naphthoquinone derivatives (compounds 3, 8, and 12) on apoptotic morphology and mitochondrial membrane potential (ΔΨm)

To investigate whether the antiproliferative effects of compounds 3, 8, and 12 were associated with apoptotic characteristics, including nuclear morphological alterations, DNA fragmentation, and DNA pyknosis. Hoechst 33342 staining was performed to characterize these features in treated HCT116 cells. Hoechst 33342 staining is a widely used method for evaluating apoptotic morphological changes because it enables the distinction between normal and apoptotic nuclei based on characteristic alterations in nuclear morphology and chromatin structure following exposure to drugs or test compounds. These apoptotic features include chromatin condensation, nuclear shrinkage, nuclear fragmentation, and the formation of apoptotic bodies, which can be visualized by fluorescence microscopy [^[43]]. As shown in Fig. 3, the morphological changes in the HCT116 cells that were pre-exposed to compounds 3, 8, and 12 (12, 35, and 7 μM, respectively) for 1 h, followed by co-culture with 1 μg/mL LPS for 48 h, exhibited significant morphological changes characteristic of apoptosis. Following 48 h, they included chromatin condensation, cell shrinkage, DNA fragmentation, and the formation of apoptotic bodies, which were comparable to those seen with the reference drug, 5-FU (25 μM). For Fig. 4, JC-1 staining revealed that compounds 3, 8, and 12 induced a loss of mitochondrial membrane potential (ΔΨm) in HCT116 cells, suggesting mitochondrial dysfunction and activation of the intrinsic apoptotic pathway. In untreated or LPS-treated cells with high ΔΨm, JC-1 accumulated within mitochondria to form J-aggregates that emitted red fluorescence, whereas cells with a reduced ΔΨm exhibited predominantly green fluorescence [^[44]]. Our results demonstrated that treatment with compounds 3, 8, and 12 for 24 h significantly reduced ΔΨm, as indicated by decreased red fluorescence and increased green fluorescence. Among the tested compounds, compound 12 produced the most pronounced increase in green fluorescence compared with the LPS-treated group and the reference drug 5-FU, indicating a greater disruption of mitochondrial function. Collectively, these findings suggest that the selected compounds differentially affect mitochondrial stability, with compound 12 exhibiting the strongest effect on mitochondrial membrane potential. The loss of ΔΨm may contribute to the induction of apoptosis and the antiproliferative effects observed in HCT116 cells. Compounds 3 and 8 also induced mitochondrial depolarization, although their effects appeared to be less pronounced than those of compound 12. These findings support the involvement of mitochondrial dysfunction in the apoptotic effects of compounds 3, 8, and 12 in HCT116 cells.### Effect of 1,4-naphthoquinone derivatives (compounds 3, 8, and 12) treatment on cell cycle distribution

In order to assess the significance of compounds 3, 8, and 12 at 12, 35, and 7 μM, respectively, on the induction of apoptosis, HCT116 cells were stained with propidium iodide (PI), and flow cytometry was used to measure DNA content and cell cycle distribution. After 1 h of pretreatment with these compounds or the reference drug, 5-FU, HCT116 cells were co-cultured with 1 μg/mL LPS for 48 h (Fig. 5). After 48 h of treatment, HCT116 cells were exposed to compound 3 (12 μM), compound 8 (35 μM), compound 12 (7 μM), or 5-FU (25 μM), corresponding to their respective IC50 concentrations. As shown in Fig. 5, the relative distribution of HCT116 cells across the different cell cycle phases revealed an increase in the sub-G1 population following treatment with compounds 3, 8, and 12. Among the tested compounds, compound 12 produced the most pronounced increase in the sub-G1 population, from 1.50 ± 0.61% in untreated control cells and 1.60 ± 0.10% in LPS-treated cells to 19.03 ± 5.25% following treatment with compound 12. In comparison, 5-FU increased the sub-G1 population to 17.40 ± 3.90% (Fig. 5A and B and Table 2). These findings suggest that compound 12 induces DNA fragmentation and apoptotic cell death in HCT116 cells, as reflected by the marked accumulation of cells in the sub-G1 population. Notably, compound 12 exhibited a stronger effect on sub-G1 accumulation than 5-FU after 48 h of treatment, suggesting a greater potential to induce apoptotic cell death under the conditions tested.### Effect of 1,4-naphthoquinone derivatives (compounds 3, 8, and 12) treatment on apoptosis

As shown in Fig. 6, flow cytometric analysis revealed that pretreatment of HCT116 cells with compounds 3, 8, and 12 at concentrations of 12, 35, and 7 μM, respectively, for 1 h, followed by exposure to LPS (1 μg/mL) for 48 h, resulted in varying degrees of apoptotic cell death. The untreated control group exhibited a low proportion of apoptotic cells, indicating a minimal basal level of apoptosis, which was comparable to that observed in the LPS-treated group. Although LPS treatment resulted in a slight increase in the apoptotic cell population, the difference was not statistically significant. Treatment with compounds 3 and 8 in the presence of LPS resulted in a moderate increase in the apoptotic cell population (Fig. 6A and B).

In contrast, the LPS/compound 12 (7 μM) treatment produced the most pronounced induction of apoptosis in HCT116 cells. Compound 12 increased the early apoptotic population to 8.40 ± 0.82% and the late apoptotic population to 46.33 ± 5.71%, compared with 8.27 ± 1.46% for the late apoptotic population in the LPS-treated control group (1 μg/mL). The marked increase in the late apoptotic population following compound 12 treatment indicates a substantial induction of apoptotic cell death. As shown in Fig. 6C, the heatmap provides a visual representation of these changes, with the strongest signal corresponding to the late apoptotic population following treatment with compound 12, whereas the increase in early apoptosis was comparatively modest. Treatment with 5-FU (25 μM), used as the reference drug, also increased apoptosis in HCT116 cells; however, the late apoptotic population was lower than that observed following treatment with compound 12. Collectively, these findings demonstrate that compound 12 effectively induces apoptotic cell death in HCT116 cells.### Binding interactions predicted by molecular docking

Molecular docking was conducted to investigate the binding interactions of the chosen compound 12, which exhibited the highest potent biological activity against human HCT116 cells. To prove the anti-inflammatory properties of compound 12, a computational docking simulation (in silico docking) of its interactions with NF-κB was performed. As shown in Fig. 7, Fig. 8, Fig. 9. Human NF-κB p50 homodimer (PDB ID: 1SVC), p50/p65 heterodimer NF-κB complex (PDB ID: 1VKX), and human IκB kinase β (PDB ID: 4KIK). The procedure for docking was successfully validated by docking the co-crystallized ligand into the NF-κB binding site using the GOLD program with the CHEMPLP scoring function. The results of molecular docking analysis demonstrated that this compound showed potential as an NF-κB inhibitor, with fitness scores similar to those of the reference drugs MLN120B and BMS-345541 (Table 3). Moreover, compound 12 could also be favorably accommodated within functionally relevant regions of the NF-κB signaling axis, including human NF-κB/p50 (Fig. 7), the p50/p65 NF-κB complex (Fig. 8), and human Ikβ kinase (Fig. 9), with binding patterns that were broadly comparable to those of the reference inhibitors MLN120B, a potent active inhibitor of IKKβ, and BMS345541, a highly selective IKK-1/IKK-2 inhibitor. In the human NF-κB/p50 model, the results showed that compound 12 was positioned near the N-terminal DNA-binding region and linker-associated surface, where it formed an interaction involving a hydrogen bond with Met208, Ser211, and Tyr60 and hydrophobic contacts with residues such as Lys147, Leu210, Lys244, Ala245, and Pro246, suggesting stable occupancy of a pocket associated with DNA recognition and transcriptional regulation. These simulations are mechanistically relevant because the p50 subunit contains two domains connected by a flexible linker, and residues in this area have been implicated in ligand recognition and modulation of NF-κB/DNA interactions [^[45]]. As shown in Fig. 8. In the p50/p65 NF-κB complex, compound 12 also occupied a binding region adjacent to the DNA-interacting interface. It established several residues interaction in active site with hydrogen bond, Halogen, van der Waals, and alkyl/π-alkyl contacts with surrounding residues, including Gly365, Val358, Gly438, His364, Ser363, Pro362, Arg356, Gly361, Val412, and Leu440, indicating that the compound may interfere with heterodimer-associated DNA binding or local conformational stabilization due to DNA binding and dimer integrity are critical for NF-κB transcriptional activity. Therefore, the interaction effect of compound 12 in this NF-κB active site, which leading to inhibitory effect on downstream inflammatory gene expression. The similarity of the binding pose of compound 12 to MLN120B and BMS345541 in this modeling study further indicates that compound 12 can bind to pharmacologically important surfaces within the NF-κB complex and may have a further inhibitory effect on inflammation [^[46]]. As shown in Fig. 9, docking against human IKβ kinase β revealed that compound 12 could also bind within the kinase-associated pocket of the upstream regulatory enzyme, with interactions involving residues such as Cys99, Gly102, Ala42, Val29, Ile165, Met96, Leu21, Glu97, Val74, and Tyr98. These contacts included conventional hydrogen bonds, carbon-hydrogen interactions, halogen interactions, and multiple hydrophobic contacts, indicating favorable compatibility between the ligand scaffold and the IKKβ binding active site. This finding importantly suggests that IKKβ is a principal upstream activator of canonical NF-κB signaling by phosphorylating IκB, leading to its degradation. Surprisingly, previous studies showed that the reference drug, MLN120B targets IKKβ to inhibit TNF-α-induced NF-κB activation [^[45]]. Therefore, the docking profile of selected compound 12 suggests that it could influence the NF-κB pathway at multiple cellular levels, potentially affecting both upstream kinase regulation and downstream transcription factor function.### Molecular dynamics study of human IκB kinase beta (IKKβ) in complex with compound 12

The binding stability of compound 12 within the kinase domain of human IKKβ was further investigated using MD simulations, with MLN120B employed as a reference inhibitor. The RMSD time profiles demonstrated lower structural deviations for compound 12 compared with MLN120B in both the ligand and the protein backbone residues located within 5 Å of the ligand (Fig. 10A). During the equilibrated 80-100 ns interval of the MD trajectories, the average RMSD values of the protein backbone were 1.87 ± 0.19 Å for the MLN120B complex and 0.64 ± 0.08 Å for the 2-chloro-3-((4-hydroxyphenyl)amino)naphthalene-1,4-dione complex, as well as the corresponding ligand RMSD values were 1.44 ± 0.47 Å and 0.48 ± 0.12 Å, respectively. These findings suggest that the complex containing compound 12 exhibited greater conformational stability, consistent with SASA and H-bond analyses. The higher structural fluctuation observed in the MLN120B complex was associated with greater solvent exposure around the binding pocket, as reflected by its higher SASA value (705 ± 75 Å2) relative to compound 12 (529 ± 48 Å2). The enhanced stability of compound 12 may be attributed to its higher #H-bonds compared with MLN120B, which was related to differences in ligand-binding orientation (Fig. 10B). In contrast, MLN120B exhibited a greater number of atomic contacts (#contacts = 235 ± 23) than compound 12 (#contacts = 209 ± 17), likely due to its larger molecular size, which provided a broader surface area for intermolecular interactions. Overall, these results suggest that compound 12 forms a more stable binding complex within the kinase domain of human IKKβ than MLN120B.

The MM/GBSA method was used to further assess the binding affinities of the protein–ligand complexes. Table 4 shows that both compounds exhibited comparable binding free energies (), with values of −29.89 ± 0.23 kcal/mol for MLN120B and −29.95 ± 0.14 kcal/mol for compound 12. The major contribution to the vacuum binding free energy () arose from van der Waals interactions (), whereas electrostatic interactions () contributed to a lesser extent, as also observed by other inhibitors in a previous study [^[47]]. This behavior is consistent with the predominance of hydrophobic hotspot residues within the binding pocket (Fig. 9) and the largely neutral characteristics of both inhibitors. Per-residue free energy decomposition () analysis identified key binding residues with stabilizing interaction energies lower than −1.0 kcal/mol (Fig. 10C, top). Both ligands shared several common hotspot residues, including Leu21, Val29, Gly102, Val152, and Ile165. Notably, compound 12 additionally exhibited significant interactions with Tyr98 and Cys99, suggesting that despite differences in chemical structure and binding geometry, compound 12 retained a binding mode similar to that of the reference inhibitor while also establishing additional stabilizing interactions with neighboring residues (Fig. 10C, bottom). The strong energy contribution of Cys99 (−3.45 kcal/mol) in the compound 12 complex can be attributed to the generation of two highly persistent H-bonds between the backbone atoms of Cys99 and the carbonyl group (98% occupancy) and −NH− group (100% occupancy) of compound 12 (Fig. 10D). These findings are consistent with the RMSD, SASA, and #H-bonds, which collectively suggested enhanced stability of the compound 12 complex. In contrast, MLN120B formed only a relatively weak H-bond with Lys106, which was not identified as a hotspot residue. Nevertheless, the larger molecular size of MLN120B and its greater #contacts enabled the complex to maintain an overall binding affinity comparable to that of compound 12.

The ATP-binding pocket located in the kinase domain of human IKKβ has been classified into several functional regions [^[48]]. The first region is the glycine-rich loop (G-loop or P-loop), which includes Leu21 and Val29. The second region corresponds to the hinge region, comprising Tyr98, Cys99, and Gly102. Val152 is located on the β-strand preceding the catalytic loop, whereas Ile165 is positioned at the beginning of the catalytic loop. Previous mutagenesis studies investigating the role of Cys99 demonstrated that the C99S mutation reduced the binding affinity of MMPP, an ATP-competitive inhibitor, toward human IKKβ [^[49]], highlighting how important this residue is for ligand recognition and stabilization. Notably, the majority of the critical binding residues found in our investigation for both MLN120B and 2-chloro-3-((4-hydroxyphenyl)amino)naphthalene-1,4-dione have also been reported in previous investigations of potential human IKKβ inhibitors, including 4-phenyl-7-azaindoles [^[50]], rotigotine [^[51]], 4-phenyl-5-p-tolyloxymethyl-4H-1,2,4-triazole thioethers [^[52]], and coumestrol [^[53]]. These observations support the reliability and consistency of the present in silico analyses with previously reported structure–activity relationships of IKKβ inhibitors. Collectively, these findings highlight compound 12 as promising inhibitors of human IKKβ.### Effect of 1,4-naphthoquinone derivatives (compounds3,8, and12) treatment inhibits TNF-α and IL-8 inflammatory cytokine release

Molecular docking analysis of compound 12 revealed favorable binding interactions with specific NF-κB target sites, involving key amino acid residues that were also engaged by the reference inhibitors MLN120B and BMS-345541. These findings suggest that compound 12 may interfere with the structural and functional activity of NF-κB-related target proteins. To further investigate the biological relevance of NF-κB-associated signaling, the secretion of inflammatory cytokines was subsequently quantified using the MILLIPLEX assay. As shown in Fig. 11, the inflammatory cytokines evaluated in this study were TNF-α (Fig. 11A) and IL-8 (Fig. 11B), both of which are widely investigated as mediators of inflammatory responses and potential targets for anti-inflammatory intervention [^[54]]. TNF-α is a key activator of the canonical NF-κB pathway and participates in a positive feedback loop that can sustain NF-κB-mediated inflammatory signaling. Upon binding to its receptor, TNF-α activates the IKK complex, leading to IκB degradation and subsequent nuclear translocation of NF-κB dimers, including the RelA/p50 complex, where they promote the transcription of genes involved in inflammation and cell survival. Therefore, the statistically significant decrease in TNF-α production following treatment with compound 12 may reflect attenuation of the inflammatory feedback mechanisms that contribute to sustained NF-κB activation in colon cancer cells [^[5]]. IL-8, an NF-κB-regulated chemokine, also plays an essential role in maintaining a pro-inflammatory tumor microenvironment by promoting immune cell recruitment, angiogenesis, and cancer cell invasion. Previous studies have reported elevated IL-8 expression in colorectal cancer tissues compared with normal mucosa and have implicated IL-8 in interactions between tumor cells and stromal and immune cells within the tumor microenvironment. Therefore, the decrease in IL-8 after treatment with compound 12 is more significant than a simple reduction in a general inflammatory marker and may reflect the capacity of these compounds to lessen the possibility of a microenvironment conducive to cancer proliferation [^[55]]. Interestingly, the results demonstrated that the synthesized chemical dramatically decreased TNF-α and IL-8 release in HCT116 cells following LPS treatment in a dose-dependent manner, compared to the control group and standard drugs. These results suggest that compound 12 may exert anti-inflammatory and anticancer effects through suppression of the NF-κB-associated cytokine network. Importantly, these findings suggest that the synthetic compound not only binds to specific enzyme sites but also exhibits biological effects consistent with the attenuation of NF-κB-associated inflammatory signaling, which may contribute to its potential anticancer activity against colorectal cancer. Although the observed decrease in TNF-α and IL-8 is consistent with suppression of inflammatory signaling, direct biochemical evidence for NF-κB or IKKβ inhibition was not obtained in the present study. Therefore, the proposed mechanism should be regarded as preliminary and requires further validation.

Additionally, compound 12 showed significant anticancer activity in our investigation, including growth suppression and apoptosis initiation in colorectal cancer cells. Moreover, our independent research in HCT116 cells demonstrated that the identical compound suppressed the expression of the inflammatory mediators TNF-α and IL-8, while also promoting apoptotic cell death, confirming our hypothesis that compound 12 could function via both anti-inflammatory and pro-apoptotic pathways. Naphthoquinone derivatives are acknowledged as redox-active molecules that can produce reactive oxygen species (ROS), a characteristic often linked to their anticancer efficacy via the promotion of oxidative stress, mitochondrial impairment, and apoptosis in cancerous cells [[^[56]], [^[57]], [^[58]]]. However, the same redox reactivity may also raise concerns regarding off-target toxicity, particularly in non-malignant cells or tissues with limited antioxidant capacity [^[59],^[60]]. Therefore, although the current results confirm the anticancer efficacy of the produced naphthoquinone derivatives, their therapeutic use should be approached with caution and not presumed to guarantee universal safety. However, due to the potential for ROS-induced cytotoxicity to create off-target effects in non-cancerous cells, additional research is necessary to elucidate precisely the mechanism of action of compound 12 and to evaluate its selectivity and safety in relevant normal-cell and in vivo models.

Antiproliferative effects of 1,4-naphthoquinone derivatives (3-18) on HCT116 cells

The cellular toxicity effects of 1,4-naphthoquinone derivatives (3-18) and the reference drug, 5-fluorouracil (5-FU) on HCT116 colorectal cancer cells and the normal African green monkey (Vero cells) cell line were determined using the MTT assay. As a result, 5-FU exhibited relatively weak cytotoxicity against HCT116 cells, while showing greater toxicity toward normal Vero cells than the corresponding 1,4-naphthoquinone derivatives. Previous studies have reported that 1,4-naphthoquinone derivatives possess anticancer activity against various cancer types, including lung, liver, breast, and colorectal cancers, while exhibiting relatively low cytotoxicity toward normal human embryonic lung MRC-5 cells [^[22]]. These findings are consistent with our experimental results. HCT116 cells were treated with the synthesized compounds (3–18) at concentrations ranging from 0 to 100 μM for 48 h, as indicated in Table 1. Among the tested compound, 2-chloro-3-(p-tolylamino)naphthalene-1,4-dione 3, 2-chloro-3-((4-nitrophenyl)amino)naphthalene-1,4-dione 8 and 2-chloro-3-((4-hydroxyphenyl)amino)naphthalene-1,4-dione 12 (Fig. 2) exhibited the strongest anticancer activity, as evidenced by their dose-dependent inhibition of HCT116 cell proliferation. The antiproliferative activity followed the order 12 > 3 > 8, corresponding to the substituent effects of R = 4-OH > 4-CH3 > 4-NO2). The IC50 value of compounds 12, 3, and 8 were 7.32 ± 0.45, 12.57 ± 2.66, and 36.54 ± 1.02 μM, respectively (Table 1). The superior activity of the compound 12 (R = 4-OH) compared with derivatives 3 (R = 4-CH3) and 8 (R = 4-NO2) can be attributed to its ability to form hydrogen-bonding interaction with amino acid residue within the target binding site. Furthermore, the hydroxyl group provides electron-donating character through resonance, which can optimize the electronic distribution of the aromatic ring and favor interactions with the receptor. The study suggests that the hydroxyl substituent provides an optimal balance of electronic modulation and hydrogen-bonding interaction, resulting in superior potency and improved selectivity. Remarkably, compound 12 emerged as the most promising candidate, exhibiting potent antiproliferative activity and greater selectivity toward HCT116 colorectal cancer cells than 5-FU, which was used as the positive pharmacological control. Compound 12 showed an IC50 value of 7.32 ± 0.45 μM against HCT116 cells, whereas 5-FU exhibited an IC50 value of 52.58 ± 1.24 μM. In contrast, most of the synthesized compounds showed relatively low cytotoxicity toward normal Vero cells, while 5-FU exhibited substantial cytotoxicity, with an IC50 value of 6.07 ± 1.22 μM. Notably, compound 12 exhibited a selectivity index (SI) of approximately 7, which was approximately 57-fold higher than that of 5-FU (SI = 0.12), indicating a markedly greater selectivity for cancer cells over normal cells. Therefore, although compound 12 demonstrated potent inhibitory activity against HCT116 cell growth, further mechanistic studies are warranted to elucidate its underlying molecular mechanisms.

Effect of 1,4-naphthoquinone derivatives (compounds 3, 8, and 12) on apoptotic morphology and mitochondrial membrane potential (ΔΨm)

To investigate whether the antiproliferative effects of compounds 3, 8, and 12 were associated with apoptotic characteristics, including nuclear morphological alterations, DNA fragmentation, and DNA pyknosis. Hoechst 33342 staining was performed to characterize these features in treated HCT116 cells. Hoechst 33342 staining is a widely used method for evaluating apoptotic morphological changes because it enables the distinction between normal and apoptotic nuclei based on characteristic alterations in nuclear morphology and chromatin structure following exposure to drugs or test compounds. These apoptotic features include chromatin condensation, nuclear shrinkage, nuclear fragmentation, and the formation of apoptotic bodies, which can be visualized by fluorescence microscopy [^[43]]. As shown in Fig. 3, the morphological changes in the HCT116 cells that were pre-exposed to compounds 3, 8, and 12 (12, 35, and 7 μM, respectively) for 1 h, followed by co-culture with 1 μg/mL LPS for 48 h, exhibited significant morphological changes characteristic of apoptosis. Following 48 h, they included chromatin condensation, cell shrinkage, DNA fragmentation, and the formation of apoptotic bodies, which were comparable to those seen with the reference drug, 5-FU (25 μM). For Fig. 4, JC-1 staining revealed that compounds 3, 8, and 12 induced a loss of mitochondrial membrane potential (ΔΨm) in HCT116 cells, suggesting mitochondrial dysfunction and activation of the intrinsic apoptotic pathway. In untreated or LPS-treated cells with high ΔΨm, JC-1 accumulated within mitochondria to form J-aggregates that emitted red fluorescence, whereas cells with a reduced ΔΨm exhibited predominantly green fluorescence [^[44]]. Our results demonstrated that treatment with compounds 3, 8, and 12 for 24 h significantly reduced ΔΨm, as indicated by decreased red fluorescence and increased green fluorescence. Among the tested compounds, compound 12 produced the most pronounced increase in green fluorescence compared with the LPS-treated group and the reference drug 5-FU, indicating a greater disruption of mitochondrial function. Collectively, these findings suggest that the selected compounds differentially affect mitochondrial stability, with compound 12 exhibiting the strongest effect on mitochondrial membrane potential. The loss of ΔΨm may contribute to the induction of apoptosis and the antiproliferative effects observed in HCT116 cells. Compounds 3 and 8 also induced mitochondrial depolarization, although their effects appeared to be less pronounced than those of compound 12. These findings support the involvement of mitochondrial dysfunction in the apoptotic effects of compounds 3, 8, and 12 in HCT116 cells.

Effect of 1,4-naphthoquinone derivatives (compounds 3, 8, and 12) treatment on cell cycle distribution

In order to assess the significance of compounds 3, 8, and 12 at 12, 35, and 7 μM, respectively, on the induction of apoptosis, HCT116 cells were stained with propidium iodide (PI), and flow cytometry was used to measure DNA content and cell cycle distribution. After 1 h of pretreatment with these compounds or the reference drug, 5-FU, HCT116 cells were co-cultured with 1 μg/mL LPS for 48 h (Fig. 5). After 48 h of treatment, HCT116 cells were exposed to compound 3 (12 μM), compound 8 (35 μM), compound 12 (7 μM), or 5-FU (25 μM), corresponding to their respective IC50 concentrations. As shown in Fig. 5, the relative distribution of HCT116 cells across the different cell cycle phases revealed an increase in the sub-G1 population following treatment with compounds 3, 8, and 12. Among the tested compounds, compound 12 produced the most pronounced increase in the sub-G1 population, from 1.50 ± 0.61% in untreated control cells and 1.60 ± 0.10% in LPS-treated cells to 19.03 ± 5.25% following treatment with compound 12. In comparison, 5-FU increased the sub-G1 population to 17.40 ± 3.90% (Fig. 5A and B and Table 2). These findings suggest that compound 12 induces DNA fragmentation and apoptotic cell death in HCT116 cells, as reflected by the marked accumulation of cells in the sub-G1 population. Notably, compound 12 exhibited a stronger effect on sub-G1 accumulation than 5-FU after 48 h of treatment, suggesting a greater potential to induce apoptotic cell death under the conditions tested.

Effect of 1,4-naphthoquinone derivatives (compounds 3, 8, and 12) treatment on apoptosis

As shown in Fig. 6, flow cytometric analysis revealed that pretreatment of HCT116 cells with compounds 3, 8, and 12 at concentrations of 12, 35, and 7 μM, respectively, for 1 h, followed by exposure to LPS (1 μg/mL) for 48 h, resulted in varying degrees of apoptotic cell death. The untreated control group exhibited a low proportion of apoptotic cells, indicating a minimal basal level of apoptosis, which was comparable to that observed in the LPS-treated group. Although LPS treatment resulted in a slight increase in the apoptotic cell population, the difference was not statistically significant. Treatment with compounds 3 and 8 in the presence of LPS resulted in a moderate increase in the apoptotic cell population (Fig. 6A and B).

In contrast, the LPS/compound 12 (7 μM) treatment produced the most pronounced induction of apoptosis in HCT116 cells. Compound 12 increased the early apoptotic population to 8.40 ± 0.82% and the late apoptotic population to 46.33 ± 5.71%, compared with 8.27 ± 1.46% for the late apoptotic population in the LPS-treated control group (1 μg/mL). The marked increase in the late apoptotic population following compound 12 treatment indicates a substantial induction of apoptotic cell death. As shown in Fig. 6C, the heatmap provides a visual representation of these changes, with the strongest signal corresponding to the late apoptotic population following treatment with compound 12, whereas the increase in early apoptosis was comparatively modest. Treatment with 5-FU (25 μM), used as the reference drug, also increased apoptosis in HCT116 cells; however, the late apoptotic population was lower than that observed following treatment with compound 12. Collectively, these findings demonstrate that compound 12 effectively induces apoptotic cell death in HCT116 cells.

Binding interactions predicted by molecular docking

Molecular docking was conducted to investigate the binding interactions of the chosen compound 12, which exhibited the highest potent biological activity against human HCT116 cells. To prove the anti-inflammatory properties of compound 12, a computational docking simulation (in silico docking) of its interactions with NF-κB was performed. As shown in Fig. 7, Fig. 8, Fig. 9. Human NF-κB p50 homodimer (PDB ID: 1SVC), p50/p65 heterodimer NF-κB complex (PDB ID: 1VKX), and human IκB kinase β (PDB ID: 4KIK). The procedure for docking was successfully validated by docking the co-crystallized ligand into the NF-κB binding site using the GOLD program with the CHEMPLP scoring function. The results of molecular docking analysis demonstrated that this compound showed potential as an NF-κB inhibitor, with fitness scores similar to those of the reference drugs MLN120B and BMS-345541 (Table 3). Moreover, compound 12 could also be favorably accommodated within functionally relevant regions of the NF-κB signaling axis, including human NF-κB/p50 (Fig. 7), the p50/p65 NF-κB complex (Fig. 8), and human Ikβ kinase (Fig. 9), with binding patterns that were broadly comparable to those of the reference inhibitors MLN120B, a potent active inhibitor of IKKβ, and BMS345541, a highly selective IKK-1/IKK-2 inhibitor. In the human NF-κB/p50 model, the results showed that compound 12 was positioned near the N-terminal DNA-binding region and linker-associated surface, where it formed an interaction involving a hydrogen bond with Met208, Ser211, and Tyr60 and hydrophobic contacts with residues such as Lys147, Leu210, Lys244, Ala245, and Pro246, suggesting stable occupancy of a pocket associated with DNA recognition and transcriptional regulation. These simulations are mechanistically relevant because the p50 subunit contains two domains connected by a flexible linker, and residues in this area have been implicated in ligand recognition and modulation of NF-κB/DNA interactions [^[45]]. As shown in Fig. 8. In the p50/p65 NF-κB complex, compound 12 also occupied a binding region adjacent to the DNA-interacting interface. It established several residues interaction in active site with hydrogen bond, Halogen, van der Waals, and alkyl/π-alkyl contacts with surrounding residues, including Gly365, Val358, Gly438, His364, Ser363, Pro362, Arg356, Gly361, Val412, and Leu440, indicating that the compound may interfere with heterodimer-associated DNA binding or local conformational stabilization due to DNA binding and dimer integrity are critical for NF-κB transcriptional activity. Therefore, the interaction effect of compound 12 in this NF-κB active site, which leading to inhibitory effect on downstream inflammatory gene expression. The similarity of the binding pose of compound 12 to MLN120B and BMS345541 in this modeling study further indicates that compound 12 can bind to pharmacologically important surfaces within the NF-κB complex and may have a further inhibitory effect on inflammation [^[46]]. As shown in Fig. 9, docking against human IKβ kinase β revealed that compound 12 could also bind within the kinase-associated pocket of the upstream regulatory enzyme, with interactions involving residues such as Cys99, Gly102, Ala42, Val29, Ile165, Met96, Leu21, Glu97, Val74, and Tyr98. These contacts included conventional hydrogen bonds, carbon-hydrogen interactions, halogen interactions, and multiple hydrophobic contacts, indicating favorable compatibility between the ligand scaffold and the IKKβ binding active site. This finding importantly suggests that IKKβ is a principal upstream activator of canonical NF-κB signaling by phosphorylating IκB, leading to its degradation. Surprisingly, previous studies showed that the reference drug, MLN120B targets IKKβ to inhibit TNF-α-induced NF-κB activation [^[45]]. Therefore, the docking profile of selected compound 12 suggests that it could influence the NF-κB pathway at multiple cellular levels, potentially affecting both upstream kinase regulation and downstream transcription factor function.

Molecular dynamics study of human IκB kinase beta (IKKβ) in complex with compound 12

The binding stability of compound 12 within the kinase domain of human IKKβ was further investigated using MD simulations, with MLN120B employed as a reference inhibitor. The RMSD time profiles demonstrated lower structural deviations for compound 12 compared with MLN120B in both the ligand and the protein backbone residues located within 5 Å of the ligand (Fig. 10A). During the equilibrated 80-100 ns interval of the MD trajectories, the average RMSD values of the protein backbone were 1.87 ± 0.19 Å for the MLN120B complex and 0.64 ± 0.08 Å for the 2-chloro-3-((4-hydroxyphenyl)amino)naphthalene-1,4-dione complex, as well as the corresponding ligand RMSD values were 1.44 ± 0.47 Å and 0.48 ± 0.12 Å, respectively. These findings suggest that the complex containing compound 12 exhibited greater conformational stability, consistent with SASA and H-bond analyses. The higher structural fluctuation observed in the MLN120B complex was associated with greater solvent exposure around the binding pocket, as reflected by its higher SASA value (705 ± 75 Å2) relative to compound 12 (529 ± 48 Å2). The enhanced stability of compound 12 may be attributed to its higher #H-bonds compared with MLN120B, which was related to differences in ligand-binding orientation (Fig. 10B). In contrast, MLN120B exhibited a greater number of atomic contacts (#contacts = 235 ± 23) than compound 12 (#contacts = 209 ± 17), likely due to its larger molecular size, which provided a broader surface area for intermolecular interactions. Overall, these results suggest that compound 12 forms a more stable binding complex within the kinase domain of human IKKβ than MLN120B.

The MM/GBSA method was used to further assess the binding affinities of the protein–ligand complexes. Table 4 shows that both compounds exhibited comparable binding free energies (), with values of −29.89 ± 0.23 kcal/mol for MLN120B and −29.95 ± 0.14 kcal/mol for compound 12. The major contribution to the vacuum binding free energy () arose from van der Waals interactions (), whereas electrostatic interactions () contributed to a lesser extent, as also observed by other inhibitors in a previous study [^[47]]. This behavior is consistent with the predominance of hydrophobic hotspot residues within the binding pocket (Fig. 9) and the largely neutral characteristics of both inhibitors. Per-residue free energy decomposition () analysis identified key binding residues with stabilizing interaction energies lower than −1.0 kcal/mol (Fig. 10C, top). Both ligands shared several common hotspot residues, including Leu21, Val29, Gly102, Val152, and Ile165. Notably, compound 12 additionally exhibited significant interactions with Tyr98 and Cys99, suggesting that despite differences in chemical structure and binding geometry, compound 12 retained a binding mode similar to that of the reference inhibitor while also establishing additional stabilizing interactions with neighboring residues (Fig. 10C, bottom). The strong energy contribution of Cys99 (−3.45 kcal/mol) in the compound 12 complex can be attributed to the generation of two highly persistent H-bonds between the backbone atoms of Cys99 and the carbonyl group (98% occupancy) and −NH− group (100% occupancy) of compound 12 (Fig. 10D). These findings are consistent with the RMSD, SASA, and #H-bonds, which collectively suggested enhanced stability of the compound 12 complex. In contrast, MLN120B formed only a relatively weak H-bond with Lys106, which was not identified as a hotspot residue. Nevertheless, the larger molecular size of MLN120B and its greater #contacts enabled the complex to maintain an overall binding affinity comparable to that of compound 12.

The ATP-binding pocket located in the kinase domain of human IKKβ has been classified into several functional regions [^[48]]. The first region is the glycine-rich loop (G-loop or P-loop), which includes Leu21 and Val29. The second region corresponds to the hinge region, comprising Tyr98, Cys99, and Gly102. Val152 is located on the β-strand preceding the catalytic loop, whereas Ile165 is positioned at the beginning of the catalytic loop. Previous mutagenesis studies investigating the role of Cys99 demonstrated that the C99S mutation reduced the binding affinity of MMPP, an ATP-competitive inhibitor, toward human IKKβ [^[49]], highlighting how important this residue is for ligand recognition and stabilization. Notably, the majority of the critical binding residues found in our investigation for both MLN120B and 2-chloro-3-((4-hydroxyphenyl)amino)naphthalene-1,4-dione have also been reported in previous investigations of potential human IKKβ inhibitors, including 4-phenyl-7-azaindoles [^[50]], rotigotine [^[51]], 4-phenyl-5-p-tolyloxymethyl-4H-1,2,4-triazole thioethers [^[52]], and coumestrol [^[53]]. These observations support the reliability and consistency of the present in silico analyses with previously reported structure–activity relationships of IKKβ inhibitors. Collectively, these findings highlight compound 12 as promising inhibitors of human IKKβ.

Effect of 1,4-naphthoquinone derivatives (compounds3,8, and12) treatment inhibits TNF-α and IL-8 inflammatory cytokine release

Molecular docking analysis of compound 12 revealed favorable binding interactions with specific NF-κB target sites, involving key amino acid residues that were also engaged by the reference inhibitors MLN120B and BMS-345541. These findings suggest that compound 12 may interfere with the structural and functional activity of NF-κB-related target proteins. To further investigate the biological relevance of NF-κB-associated signaling, the secretion of inflammatory cytokines was subsequently quantified using the MILLIPLEX assay. As shown in Fig. 11, the inflammatory cytokines evaluated in this study were TNF-α (Fig. 11A) and IL-8 (Fig. 11B), both of which are widely investigated as mediators of inflammatory responses and potential targets for anti-inflammatory intervention [^[54]]. TNF-α is a key activator of the canonical NF-κB pathway and participates in a positive feedback loop that can sustain NF-κB-mediated inflammatory signaling. Upon binding to its receptor, TNF-α activates the IKK complex, leading to IκB degradation and subsequent nuclear translocation of NF-κB dimers, including the RelA/p50 complex, where they promote the transcription of genes involved in inflammation and cell survival. Therefore, the statistically significant decrease in TNF-α production following treatment with compound 12 may reflect attenuation of the inflammatory feedback mechanisms that contribute to sustained NF-κB activation in colon cancer cells [^[5]]. IL-8, an NF-κB-regulated chemokine, also plays an essential role in maintaining a pro-inflammatory tumor microenvironment by promoting immune cell recruitment, angiogenesis, and cancer cell invasion. Previous studies have reported elevated IL-8 expression in colorectal cancer tissues compared with normal mucosa and have implicated IL-8 in interactions between tumor cells and stromal and immune cells within the tumor microenvironment. Therefore, the decrease in IL-8 after treatment with compound 12 is more significant than a simple reduction in a general inflammatory marker and may reflect the capacity of these compounds to lessen the possibility of a microenvironment conducive to cancer proliferation [^[55]]. Interestingly, the results demonstrated that the synthesized chemical dramatically decreased TNF-α and IL-8 release in HCT116 cells following LPS treatment in a dose-dependent manner, compared to the control group and standard drugs. These results suggest that compound 12 may exert anti-inflammatory and anticancer effects through suppression of the NF-κB-associated cytokine network. Importantly, these findings suggest that the synthetic compound not only binds to specific enzyme sites but also exhibits biological effects consistent with the attenuation of NF-κB-associated inflammatory signaling, which may contribute to its potential anticancer activity against colorectal cancer. Although the observed decrease in TNF-α and IL-8 is consistent with suppression of inflammatory signaling, direct biochemical evidence for NF-κB or IKKβ inhibition was not obtained in the present study. Therefore, the proposed mechanism should be regarded as preliminary and requires further validation.

Additionally, compound 12 showed significant anticancer activity in our investigation, including growth suppression and apoptosis initiation in colorectal cancer cells. Moreover, our independent research in HCT116 cells demonstrated that the identical compound suppressed the expression of the inflammatory mediators TNF-α and IL-8, while also promoting apoptotic cell death, confirming our hypothesis that compound 12 could function via both anti-inflammatory and pro-apoptotic pathways. Naphthoquinone derivatives are acknowledged as redox-active molecules that can produce reactive oxygen species (ROS), a characteristic often linked to their anticancer efficacy via the promotion of oxidative stress, mitochondrial impairment, and apoptosis in cancerous cells [[^[56]], [^[57]], [^[58]]]. However, the same redox reactivity may also raise concerns regarding off-target toxicity, particularly in non-malignant cells or tissues with limited antioxidant capacity [^[59],^[60]]. Therefore, although the current results confirm the anticancer efficacy of the produced naphthoquinone derivatives, their therapeutic use should be approached with caution and not presumed to guarantee universal safety. However, due to the potential for ROS-induced cytotoxicity to create off-target effects in non-cancerous cells, additional research is necessary to elucidate precisely the mechanism of action of compound 12 and to evaluate its selectivity and safety in relevant normal-cell and in vivo models.

Conclusions

In conclusion, the synthesized anilino-1,4-naphthoquinone derivatives, particularly compound 12, demonstrated promising anticancer activity against colorectal cancer cells through anti-inflammatory and pro-apoptotic effects. Molecular docking analysis suggested favorable interactions of compound 12 with NF-κB-related targets, including NF-κB/p50, the p50/p65 complex, and IKKβ. Furthermore, compound 12 reduced TNF-α and IL-8 production, supporting its potential to attenuate NF-κB-associated inflammatory signaling. Compound 12 also promoted apoptotic cell death in HCT116 colorectal cancer cells, accompanied by mitochondrial membrane depolarization and an increase in the sub-G1 population. Collectively, these findings highlight compound 12 as a promising lead compound for further investigation and development as a potential therapeutic candidate for colorectal cancer.

CRediT authorship contribution statement

Sakdiphong Punpai: Writing – review & editing, Writing – original draft, Visualization, Validation, Resources, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization, Project administration. Panupong Mahalapbutr: Writing – review & editing, Writing – original draft, Visualization, Software. Panyakorn Taweechat: Writing – review & editing, Writing – original draft, Visualization, Software. Ronnakorn Leechaisit: Methodology, Investigation. Kiattawee Choowongkomon: Writing – review & editing, Supervision, Software. Ratchanok Pingaew: Writing – review & editing, Writing – original draft, Supervision, Resources. Wanlaya Tanechpongtamb: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Funding acquisition, Formal analysis, Data curation, Conceptualization.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.