Unveiling the anticancer effect of cytisine against oral cancer: Insights from bioinformatics and in-vitro approaches

Abstract

Introduction:

Oral squamous cell carcinoma (OSCC), the most common malignancy, involves molecular alterations including p53 mutations, TGF-α, and EGFR overexpression. Targeting biomarkers is essential for effective therapy. Cytisine, a natural smoking-cessation alkaloid with diverse pharmacological activities, was evaluated using bioinformatics to identify potential anticancer targets and subsequently validated in vitro.

Materials and Methods:

The in silico analysis included predictions of toxicity, drug-likeness, and potential targets for cytisine. OSCC-associated targets were retrieved from GeneCards and ACTREC databases. Cytoscape was used to construct cytisine-target pathways. Molecular docking and 100 ns molecular dynamics simulations were performed to evaluate cytisine binding affinity and structural stability to key targets. The in vitro validation was performed using the MTT assay and cell migration assay on an epithelial carcinoma cell model (KB-3-1 cell line).

Results:

Cytisine was non-toxic and modulated 23 target proteins and 13 signalling pathways. Gene enrichment analysis revealed significant interactions with EGFR and TP53. Molecular docking and 100 ns molecular dynamics simulation showed strong intermolecular interactions and low binding energies with EGFR (−6.9 kcal/mol) and TP53 (−7.7 kcal/mol) compared with standard drugs. In vitro assays demonstrated significant cytotoxicity and inhibition of cell migration in epithelial carcinoma cell lines while maintaining the viability of L929 normal fibroblast cells.

Conclusion:

Cytisine exhibits strong interactions and lower binding energies with EGFR and TP53, and cytotoxic and anti-migratory effects against the epithelial carcinoma cell model. However, further validation using primary oral squamous carcinoma cell lines and in vivo studies is necessary to confirm its therapeutic relevance to epithelial carcinoma.

INTRODUCTION

Oral cancer, being the most common malignancy, stands in the 16^th rank globally in terms of cancer incidence. Asia accounts for the highest prevalence of lip and oral cavity cancers.[1] It is frequently known to develop through oral potentially malignant disorders (OPMDs). Although these lesions are easily detectable on routine examination, they are unfortunately diagnosed only at later stages, when they have progressed to cancer. The reasons are mainly due to patient ignorance and stigma surrounding seeking dental care at an early stage.[2]

Over the years, several biological markers have been identified and described in tumours of the head and neck. Research has indicated that p53 mutations and overexpression of markers such as transforming growth factor-α (TGF-α), epidermal growth factor receptor (EGFR), and cyclin D1 are linked to cancer. Amongst these, the molecules that are prominent in their roles in cancer progression and as critical mediators of signal transduction pathways are EGFR and p53.[3,4]

EGFR is a superficial tyrosine kinase receptor protein that plays a key role in embryogenesis and also plays a significant role in the suppression of apoptosis, spread, invasion, and metastasis of tumours. It has been speculated that EGFR is overexpressed in 80% of OSCCs and is associated with a more invasive phenotype, unfavourable prognosis, and therapeutic resistance.[5] Hence, therapeutic agents targeting EGFR have opened new possibilities for the treatment of oral carcinoma. Such agents function by blocking EGFR activation and its associated regulatory pathways, thereby hindering tumour development and metastasis. Studies have shown that EGFR-targeted therapies are effective in epithelial carcinoma, such as OSCC, both when used as monoclonal therapy or in combination approaches.[6,7,8]

The guardian of the genome, p53, primarily functions as a tumour-suppressing transcription factor. In OSCC, p53 gene mutation is frequently induced by upregulation of transcription factors in response to tobacco-related chemical and physical agents. This results in dysregulation of the cell cycle, primarily at the G1/S phase, ultimately leading to the emergence of neoplastic traits.[9] p53-targeted therapies have proven promising for a wide range of human tumours, as 50% of tumours carry mutant p53. This approach focuses on compounds that can either restore wild-type p53 function or eliminate the mutant form.[10,11]

Tumour suppressor genes and oncogenes are crucial drivers in the pathogenesis of oral carcinoma, emphasizing its importance of targeted therapy. It is crucial to understand molecular changes in tumour cells and to identify reliable biomarkers to develop more effective targeted treatments. Hence, identifying a promising compound or drug and formulating a therapeutic strategy is essential for improving clinical outcomes in OSCC.[12,13,14]

Cytisine is a naturally occurring plant alkaloid with bioactive properties. In European countries, it has been utilised as an effective treatment for smoking cessation. Additionally, cytisine has demonstrated benefits in reducing alcohol consumption, protecting cardiovascular health, regulating blood sugar, providing neuroprotection, preventing and treating osteoporosis, and exhibiting anti-tumour properties.[15,16] Research through in vitro studies has validated its anti-cancer effects in various systemic cancer cell lines. Despite its potential, cytisine remains an “unmapped” alkaloid in the context of OSCC, with limited bioinformatic data available to elucidate its precise molecular targets and mechanisms of action. Hence, this study aimed to map the interaction of cytisine with other molecules and their binding affinity related to oral cancer through bioinformatics and further to validate its anti-cancer activity through cytotoxicity assay and migration assay against epithelial carcinoma model and cytoprotective effect on normal fibroblast cell lines.

MATERIAL AND METHODS

Bioinformatics protocol

he study was conducted after obtaining the ethical clearance from ethical board of university (KAHER/EC/24-25/365) and IRB (271). Standard bioinformatics protocols were used to extract data from specific databases. The canonical “SMILES (Simplified Molecular Input Line Entry System)” notation for cytisine, representing its molecular structure in text format, was obtained from the PubChem database (https://pubchem.ncbi.nlm.nih.gov/) and utilised as the molecular input for subsequent analyses.

Toxicity profiling of cytisine

We employed PKCSM, ADVER pred and ADMETSR lab 2.0 (https://admetmesh.scbdd.com/) computational tools to assess the toxicity profile and pharmacokinetics of cytisine.

Assessment of drug likeness score

MolSoft was utilised to evaluate and compute the drug-likeness score of cytisine. We analysed various molecular descriptors that are commonly associated with drug-like behaviour. Information on the number of hydrogen bond donors and acceptors, and the molecular weight of cytisine was obtained.

Target prediction of cytisine

Using the Super-Pred server, the protein target of cytisine was predicted (https://prediction.charite.de/).

Retrieval of oral cancer targets

A literature review was conducted to retrieve the potential

targets related to oral cancer from two prominent databases: GeneCards (https://www.genecards.org/) and the Oral Cancer Database rom the Advanced Centre for Treatment, Research and Education in Cancer (ACTREC)” (https://actrec.gov.in/OCDB/index.htm). On searching both databases, the target genes were obtained after deduplication. The retrieved data were integrated, and genes listed in both Gene Cards and the Oral Cancer Database from ACTREC were considered high-priority targets for further investigation.

Gene enrichment of regulated targets

To explore the pathways modulated by a panel of genes, protein targets were analysed using the STRING database (https://string-db.org/). Additionally, pathway information from “Kyoto Encyclopaedia of Genes (KEGG) pathway” (https://www.genome.jp/kegg/pathwayy.html) and relevant scientific literature was used to construct a comprehensive gene involved in the regulation of pathways linked to oral cancer.

Network construction

A network connecting cytisine, its predicted targets, along with its associated biological pathways, was constructed using Cyto-scape (https://cytoscape.org/). Their interactions within the network were evaluated using the edge count topological metric. The node sizes were scaled according to the lower value, and the node colours were mapped from lower values to more striking colours to enhance clarity and visual impact.

Molecular docking

Following the network analysis, molecular docking studies were conducted on cytisine targeting EGFR and TP53, which were identified as key proteins within the network. The docking process was conducted using AutoDock Vina, executed through the POAP pipeline with the docking grid defined around the active site. BIOVIA Discovery Studio Visualizer 2019 was utilised to assess binding energies and visualise molecular interactions. Erlotinib and PC14586 were employed as standard drugs for EGFR and TP53, respectively.

Molecular dynamics

GROMACS ver 2013 software programme was used to conduct MD simulations. All atom molecular dynamics simulations were conducted for 100 ns in a solvent explicit model to evaluate their intermolecular interactions and conformational stability of the most stable protein-ligand complexes.

Statistical analysis

For network modelling interpretation, the number of edges was defined. Binding affinities were computed as kcal/mol. RMSD was employed to define structural integrity, and RMSF was used for the assessment of fluctuations found by dynamic simulation. The cell culture assays were assessed through viability percentage for MTT, and an independent t-test to compare the wound area remaining after 24 hours between treated and untreated groups.

In vitro cytotoxicity assay protocol

Epithelial carcinoma cell line model (KB-3-1) and mouse embryonic fibroblast cells (L929) were procured from the National Centre for Cell Sciences, Pune, India. Cytisine was purchased from PSM Scientific Systems, India.

MTT (3-(4, 5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide) Assay

A seeding density of 5 × 10⁵ cells/well was used for the epithelial carcinoma cell line (KB3-1), and 10 × 10⁵ cells/well for L929 cells, in a 96-well plate with a final volume of 100 μl per well. The cells were incubated in a humidifying chamber at 37°C with 5% CO₂ for 24 hours. Cells were treated with six serial dilutions of cytisine from 500 μg/ml to 15.625 μg/ml for 24 hours. Untreated cells served as negative controls. Following 24 hours of incubation, the cytisine compound was removed, the culture was replaced with fresh complete media, and the culture was incubated for an additional 24 hours to allow for cell recovery. Subsequently, 20 μL of MTT reagent was added to each well, and the wells were incubated at 37°C for four hours. 100 μL of dimethyl sulfoxide (DMSO) was added to solubilise the formazan crystals. Following the addition of 10 μL of glycine buffer to all wells to optimise the pH, absorbance was recorded at 570 nm using a microplate reader. The cell viability inhibition percentage was calculated as Viability inhibition = [(optical density of control cellsoptical density of treated cells)/optical density of control cells] ×100. The IC₅₀ value was calculated at all the concentrations of cytisine for both cell lines.

Cell migration assay

Six-well plates were seeded with epithelial carcinoma cell lines and incubated to allow monolayer formation. The migration assay was set up with a negative control without the compound and the test compound (IC₅₀: cytisine). Once the monolayer of cells was formed, a scratch or wound was created using a sterile microtip (200 μl). Subsequently, a gentle phosphate-buffered saline (PBS) wash was given to remove the dislodged cells. The cells were exposed to 406.55 μg/ml (IC₅₀) of cytisine diluted in media, and fresh complete media was added and incubated for 24 hours. The images were captured focusing on the wound area, starting from 0 hours and every 4^th hour (0, 4, 8, 12, 16, 20, 24 hours) until the 24^th hour, until the wound is approximated in the negative control.

The area of the wound was analysed using “ImageJ software bundled with Zulu OpenJDK 13.0.6” using the wound healing size tool plugin. The areas of the wound were incorporated in an Excel sheet, and the migration percentage was calculated using a formula. The percentage of migration was calculated by area of wound at zero hour minus area of wound at every fourth hour divided by area of wound at zero hour multiplied by 100.

RESULTS

Toxicity profile and drug likeliness of cytisine

Cytisine showed a non-toxic profile. The predicted molecular weight of cytisine was 190.11, having 2 hydrogen bond acceptors, 1 hydrogen bond donor, and 1.01 MolLogP. Permeation of the blood-brain barrier with cytisine was evident.

Target prediction and retrieval of oral cancer genes

We retrieved 6489 oral cancer-inducing genes from the GeneCard database and 375 genes from the oral cancer database from ACTREC. We filtered for common genes across both databases and ultimately retrieved 291 genes.

Enrichment analysis and network construction of regulated targets

Gene enrichment analysis revealed that cytisine was predicted to target 23 protein targets, with the highest predictions for EGFR and TP53, and to control 13 pathways in the KEGG database [Figure 1]. Finally, the network linking compound-associated proteins and pathways was built using Cytoscape ver 3.6.1 and assessed using topological parameters, including edge count [Figure 2].

Figure 1.

The protein-protein interactions of cytisine targets using the string database

Figure 2.

Network representation of cytisine anti-cancer activity pathways with their probable protein molecules involved in oral squamous cell carcinoma using Cytoscape software

Molecular docking

The results showed that the EGFR complex has the slowest binding energy for cytisine with -6.9 kcal/mol via forming 1 hydrogen bond and 6 non-hydrogen bond interactions (Leu149 (2), Ala48 (2), Val31 (2), respectively, when compared to Erlotinib with a binding affinity of -6.1 kcal/mol, which formed 2 hydrogen bonds and 7 non-hydrogen bonds (Ala27, Val31, Asn147, Ala48, Leu97 (2), Leu23) [Table 1]. All the interactions involved were formed at the active site residues. Similarly, TP53 complex showed binding energy of -7.7 kcal/mol, which formed 1 H- bond and 3 non-hydrogen bonds, namely (Cys242 (2), Cys176) compared to standard with binding affinity of -5.9 kcal/mol with 2 hydrogen bonds and 7 non-hydrogen bonds like (His179, Tyr239, Cys242 (2), Pro177, Met423 (2)), respectively [Table 2].

Table 1.

Binding energy of cytisine-erlotinib with EGFR complex

Complex	Binding energy (kcal/mol)	*HBI	*N-HBI	Total number of interactions
Cytisine	-6.9	Thr95…=O	Leu149 (2), Ala48 (2),Val31 (2)	7
Erlotinib	-6.1	Asn147…O, Thr159…O	Ala27,Val31,Asn147, Ala48,Leu97 (2),Leu23	9

*HBI: Hydrogen bond interactions, *NBI: Non-hydrogen bond interactions

Table 2.

Binding energy of cytisine-PC14586 with TP53 complex

Complex	Binding energy kcal/mol	*HBI	*N-HBI	Total number of interactions
Cytisine	-7.7	Met243	Cys242 (2), Cys176	4
PC14586	-5.9	His179…N, Ser182…N	His179, Tyr239, Cys242 (2), Pro177, Met423 (2)	9

*HBI: Hydrogen bond interactions, *NBI: Non-hydrogen bond interactions

Molecular dynamics simulations

The complexes, cytisine with EGFR and TP53, along with standard drugs Erlotinib and PC14586, were subjected to a 100 ns MD simulation. To gain better insights into structural stability, RMSD, RMSF, radius of gyration, SASA, and hydrogen bond interactions were analysed.

Root mean square deviation (RMSD) analysis

The backbone RMSD values across the trajectories demonstrated steady dynamics for all of the simulated complexes over the 100 ns MD simulation. The stable dynamics were observed with the complex cytisine-EGFR and erlotinib-EGFR during the 100 ns run. The average backbone RMSD values for cytisine and erlotinib were found to be 0.89 and 0.93 Å and the complex RMSD values were 1.21 and 1.67 Å, respectively, after the 20 ns equilibration period. The complex cytisine-p53 showed a steady dynamic, whereas PC14586-p53 was found to escape the binding pocket after 20 ns. Similarly, the average backbone RMSD for cytisine and PC14586 were found to be 1.17 and 1.73 Å, and the complex RMSD were 0.71 and 2.67 Å, respectively [Figure 3]. Furthermore, RMSD fluctuation for ligand-bound complexes was analysed. EGFR followed the trend similar to that of backbone RMSD, but for Tp53, it showed significant fluctuation.

Figure 3.

Root mean square deviation plot for backbone atoms and complex of cytisine with erlotinib (standard drug for EGFR) and PC14586 (standard drug for p53) over 100 ns MD simulation. The average backbone RMSD values for cytisine-erlotinib were 0.89 and 0.93 Å and the average backbone RMSD for cytisine-PC14586 were 1.17 and 1.73 Å. (a). The complex RMSD for cytisine-erlotinib were 1.21 and 1.67 Å, the RMSD for cytisine-PC14586 were 0.71 and 2.67 Å (b)

Root mean square fluctuation (RMSF) analysis

The C- terminal showed the maximum peak of 7.6 Å, and multiple peaks were also observed at 3.85, 2.87 and 4.6 Å with aa20 to aa55 and aa285 to aa305. The C- terminal showed the maximum peak of 7.6 Å, and multiple peaks were also observed at 5.95, 3.37 and 4.16 Å with aa20 to aa50 and aa90 to aa150 [Figure 4 a and b].

Figure 4.

Root mean square fluctuation plot for C-alpha atoms of cytisine, EGFR and p53 complexes over 100 ns MD simulation. Most residues exhibited fluctuations below 2 Å, while higher peaks were observed in loop and terminal regions in EGFR–cytisine (7.6 Å) and in p53–cytisine (7.6 Å). (a-b). Compactness measure (radius of gyration) of cytisine with EGFR and TP53 plots demonstrating stable structural compactness throughout the simulation, with EGFR complexes maintaining Rg values around 2.08–2.12 Å and TP53 complexes between 1.52 and 1.68 Å.(c-d). Solvent accessible surface area for cytisine with EGFR and TP53, indicating stable solvent exposure patterns, where EGFR complexes ranged from ~ 101.6 to 111.7 nm² and TP53 complexes ranged from 160.6 to 176.3 nm². Overall, cytisine-bound complexes exhibited stable conformational behaviour comparable to or better than the reference ligands erlotinib and PC14586 (e-f)

Radius of Gyration (Rg) values and Solvent accessible surface area (SASA)

To evaluate the structural compactness and folding of protein-ligand complexes, we analysed the Rg, which reflects the spatial distribution of the protein’s backbone atoms. EGFR standard complex and cytisine complex showed a gradual decrease in Rg value, which were found to be 2.08 and 2.12 Å. Similarly, the TP53 standard complex showed a decrease in Rg values, which were found to be 1.68 and 1.52 Å [Figure 4 c and d]. SASA value occupied by EGFR complexes ranges between 101.6 and 111.7 nm², whereas SASA value occupied by TP53 complexes ranges between 160.6 and 176.3 nm² [Figure 4 e and f]. The Rg values for both the standard complexes (Erlotinib and PC 14586) and the cytisine complexes remained consistently confined within a narrow range. Additionally, the values of SASA of all complexes indicate structural stability, suggesting the compact globular conformation throughout the 100 ns simulation.

Hydrogen bond interactions in simulated complexes

Hydrogen bonds play a crucial role in maintaining the stability of protein-ligand complexes. Therefore, the formation and persistence of hydrogen bonds were monitored during the 100 ns simulation. Erlotinib complex showed 4 H-bonds with 3 consistent bonds up to 100 ns, cytisine complex also formed 4 H-bonds with partial presence till 100 ns run [Figure 5 a and b]. PC14856 complex formed 5 H-bonds with none consistent up to 100 ns, cytisine complex also formed 4 H-bonds with 2 consistent till 100 ns run [Figure 5 c and d].

Figure 5.

Hydrogen bond interaction formed between erlotinib and cytisine complex during 100 ns MD simulation (a and b). A hydrogen bond interaction formed between PC14586 and the cytisine complex during 100 ns MD simulation (c and d)

Principal component of analysis

The PCA analysis shows the conformation changes per residue of motion within the binding pocket over the ligand. In complex EGFR-cystine, a lesser conformation and evolution of clusters ranging from +4 to -4 when compared to the standard erlotinib, with an eigen vector of 2 to -2. Similarly, the eigenvalue for p53 complexed with cytisine and standard inhibitor, the values were observed to range from +2 to -3 and +3 to -3, posing higher structural diversity and significant changes compared to cytisine. Therefore, the generated value shows cytisine is well stabilised and shows lesser conformation change throughout the course of simulation [Figure 6].

Figure 6.

Principal component analysis of protein-ligand complexes. The EGFR-Erlotinib complex appears more stable than EGFR-cytisine, due to tighter clustering in PCA and sharper eigenvalue drop-off (a). p53-cytisine complex is potentially more localised and stable or less flexible compared to p53-PC14586 (b)

Dynamic cross-correlation matrix (DCCM) analysis

The collective motion of the ligand-binding domains was analysed using DCCM, showing the 2D matrix representing the correlation in the residue motion in a 100 ns run. The positive regions, coloured amber, represent strongly correlated motions, whereas the negative regions, which are coloured blue, represent anticorrelated regions. The EGFR-cytisine occupied the maximum region, showing a positive correlation mainly at the binding pocket region compared to the erlotinib, revealing opposite movement at both ends. Interestingly, we noticed a positive correlation in p53 upon binding to cytisine and the standard. Thus, DCCM favours the stable complex formation during the simulation [Figure 7].

Figure 7.

Dynamic cross-correlation matrix of Ca⁺ atoms observed in complexes for EGFR (a), TP53 (b), with cytisine. The positive regions, which are coloured amber, represent strongly correlated motions of Ca atoms, whereas the negative regions, which are coloured blue, represent anticorrelated motions

Cytotoxicity and cytoprotective assay

At 500 μg/ml, the majority of the epithelial carcinoma model cells showed round and granular morphology suggestive of cell death. However, in subsequent concentrations, only a few of the cells showed altered morphology. On the contrary, L929 cells showed minimal change in the morphology and maintained viability at all the concentrations of cytisine [Figure 8]. The half maximal inhibitory concentration (IC₅₀) of cytisine for epithelial carcinoma model cells was 406.55 μg/ml, and L929 cells was 1178.68 μg/ml.

Figure 8.

Epithelial carcinoma model and L929 (mouse embryonic fibroblast) cell lines. The untreated (a) and treated (b) cytisine compound on L929 cells showed similar morphology with good viability of cells. The epithelial carcinoma cell model, untreated cells maintained the morphology of cells (c), and cytisine treated cells showed a change in morphology to round cells, membrane blebbing and granular (d). The graphs e and f represent the viability percentage with respective IC₅₀ values of the epithelial carcinoma cell model and L929 cell lines

Cell migration assay

The migration assay revealed that at zero hour, the area percentage of wound in untreated epithelial carcinoma model cells was 56%, and at the end of 24^th hour, the area remaining was only 8% with 85% of migration of cells. However, in cells treated with cytisine, the area percentage of the wound was 50% at zero hour, and at the end of 24^th hour, the area remaining was 21%, suggesting 57% of migration of cells [Figure 9]. The mean difference between treated and untreated groups was 12.64 with a P value of 0.0063, indicating that the difference between the groups was statistically significant [Table 3]. Thus, on comparing with untreated cells, the compound cytisine showed significant inhibition of the migration of the epithelial carcinoma model.

Figure 9.

Cell migration assay on oral cancer cell lines. The untreated oral cancer cells showed 85% closure of the wound area from zero hours to 24 hours (a-d). The cytisine-treated oral cancer cells showed inhibition of migration of 57% of oral cancer cells at 24 hours (e-h)

Table 3.

Effect of cytisine on migration through wound healing assay comparing the treated cells and untreated cells at the end of 24 hours, by an independent t-test

Wound area remaining at 24 hours
Group	Mean	SD	Mean Diff	P
Treated	21.273	3.1	12.642	0.0063*
Untreated	8.631	2.8
Wound closure percentage at 24 hours
Treated	56.93	5.2	27.5	0.0033*
Untreated	84.43	5.5

*P < 0.05 significant, SD: standard deviation

DISCUSSION

Tumorigenesis is an intricate process wherein normal cells undergo malignant transformation, driven by genetic alterations, environmental exposure and lifestyle changes. Mutations in key genes such as EGFR, P53 and HER2 play a significant role in the onset and advancement of oral cancer.[17] Specific mutations serve as potential biomarkers for the diagnosis and prognosis of oral cancer. They also guide the selection of targeted therapies, highlighting the crucial role of molecular diagnostics in personalised treatments.[18,19] However, despite early therapeutic success in targeting EGFR and TP53, significant hurdles persist, such as the development of resistance and associated toxicities.

Cytisine is a plant alkaloid commonly known as a smoking cessation aid. The additional effect of cytisine as an anti-cancer agent has been supported by in vitro assays in hepatoma, lung cancer, pharyngeal carcinoma, and breast adenocarcinoma cell lines. But the effect of cytisine on oral cancer has not been proven. Thus, to explore the anticancer property of cytisine, we first studied it using bioinformatics and then further validated it in an epithelial carcinoma model (KB-3-1 cells).

The bioinformatic analysis proved that compound cytisine showed a non-toxic profile and targeted 23 proteins, namely EGFR, TP53, GSK3B, ABL1, FYN, STAT3, SERPINE1, ZAP70, etc., and modulated 13 pathways for carcinogenesis, including apoptosis, PI3K-Akt signalling pathway, mTOR signalling pathway, cell cycle, p53 signalling pathway and several other pathways. Since EGFR and TP53 were highly projected protein targets, we docked cytisine with EGFR and TP53 to identify the binding affinity of cytisine with EGFR and TP53 to assess the ability of cytisine as an anticancer agent. Our docking analysis indicated that cytisine effectively binds to the EGFR’s ATP-binding pocket, with a binding score of -6.9 kcal/mol, forming stabilising hydrogen bonds and hydrophobic interactions compared to the standard drug erlotinib with a binding score of -6.1 kcal/mol. This interaction suggests a mechanism through which cytisine may inhibit EGFR phosphorylation, thereby reducing cell proliferation. There are various studies regarding the binding affinity of natural compounds targeting EGFR and TP53. Park et al.[20] found that the inhibition mechanism of the EGFR inhibitor erlotinib blocks EGFR signalling by competitively binding at the ATP site. Singh et al.[21] studied the docking of natural compounds such as Quercitin and epigallocatechin gallate for EGFR inhibition, showing better binding energies of -6 to -7.6 kcal/mol. Another study on docking EGFR with Triphala showed binding energy ranging from -6.3 to- 6.6 kcal/mol.[22]

Thus, cytisine has shown consistent results compared with previous studies, with a binding energy of -6.9 kcal/mol, suggesting that cytisine has strong binding affinity for EGFR. Docking analysis of TP53 revealed that cytisine binds effectively to the DNA-binding domain of TP53, forming stabilising hydrogen bonds and van der Waals interactions that may support TP53 structural integrity. This stabilisation suggests cytisine could potentially restore or enhance TP53 function. Molecular docking of cytisine with TP53 showed the slowest binding energy of -7.7 kcal/mol, while the standard drug PC14586 showed the score of -5.9 kcal/mol, which proves that effective binding is present between cytisine and TP53. Our study showed stronger binding affinity than a previous study on Moringa Oleifera targeting TP53, which reported binding energies ranging from -5.29 to -6.72 kcal/mol.[23] Another study on natural compounds such as quercetin, podophyllotoxin, and rutin against p53 reported binding energies of −7.3 and −6.8 kcal/mol, which were in accordance with our findings.[24] Compared with previous literature, cytisine has shown a binding energy of -7.7 kcal/mol with TP53, which is much lower than those reported in other studies, suggesting significant binding with TP53.

Ideally, a lower binding energy suggests stronger binding interactions. Cytisine demonstrated favourable binding affinities to both EGFR and TP53. Hence, it is tempting to speculate that cytisine exhibits strong binding affinities for both EGFR and TP53 compared with other phytochemicals investigated for these two targets.

The molecular dynamics simulation gave us insights into the structural stability for a 100-ns run. The MD simulation results demonstrate that cytisine remains stably bound to both EGFR’s ATP-binding pocket and TP53’s DNA-binding domain over the course of the simulation. Root mean square deviation (RMSD) analysis shows relatively stable values for the protein backbones and ligand conformations in both complexes, indicating that cytisine binding does not induce significant destabilisation of EGFR or TP53. The reference ligand PC14586 showed displacement from the TP53 binding pocket after around 20 ns during the MD simulation, indicating decreased complex stability under the simulated conditions. The observed behaviour may be due to inherent conformational dynamics rather than methodological bias because both ligand systems were subjected to the same simulation settings, such as force field, solvation environment, equilibration technique, and simulation time. However, this observation should be regarded with caution as it constitutes a limitation of the comparison research.

RMSF values remained below 7, indicating the stability during the simulation. Furthermore, the Rg values were consistent, suggesting a stable structure throughout the simulation. The SASA values align with Rg values, predicting minimal confrontational changes during the simulation run. Cytisine ligand complexes showed stability with no major alterations in the protein structure throughout the 100 ns simulation. Overall, cytisine has been shown to possess anticancer properties by acting on multiple targets, mainly EGFR and Tp53.

Furthermore, the cytisine anti-cancer activity was validated through cytotoxicity against epithelial cancer cell line lines and cytoprotective against normal fibroblasts (L929). Over the years, researchers have demonstrated the anticancer activity of cytisine on various cell lines. Yu L et al.[25] observed dose-dependent suppression of HepG2 cell growth with cytisine IC₅₀ value of 5.36 mmol·l^-1. It also induced endoplasmic reticulum stress in HepG2 cells.[26] Peng et al.[27] assessed cytotoxicity of cytisine against breast cancer cell lines, proving significant anticancer activity of cytisine exhibiting an IC₅₀ value of 19.2 μM. Cytisine is also known to inhibit the growth of lung cancer cell lines such as NCI-H23, A549, and NCI-H460. The study demonstrated that cytisine reduced cell viability in a dose-dependent manner, with IC50 values ranging from 26.83 to 32.45 μM, suggesting its potential to suppress lung cancer cell proliferation.[28] In assessing the cytotoxicity of cytisine in an epithelial carcinoma cell model and determining its cytoprotective effect on fibroblasts, we observed that the viability of epithelial carcinoma cells was dose-dependently inhibited following exposure to cytisine, while the viability of L929 cells was maintained at all concentrations. The IC₅₀ value for cytisine was 406.55 μg/ml, which is higher than that of traditional anticancer drugs, which show activity at lower concentrations. This could be due to cytisine being an isolated compound, with a preliminary mechanism of action involving cessation of the smoking habit through the nicotinic Acetylcholine receptor pathway. Moreover, the normal fibroblast cell line (L929), used as a control, maintained a high viability percentage across all concentrations, suggesting that cytisine is cytoprotective. In normal cells, antioxidants regulate intracellular ROS levels and maintain a concentration that supports important cellular processes such as cell division, proliferation, and differentiation. While normal cells can tolerate moderate ROS levels required for survival, cancer cells become vulnerable when ROS accumulation exceeds the threshold, ultimately resulting in cell death.[29,30] This may represent a potential mechanism for cytoprotection in normal cells.

Following the IC50 value, the cell migration assay was performed to assess the potential of cytisine on cell migration. We found inhibition of migration of cytisine-treated cells, with only 57% of migration compared to untreated cells that showed 85% of migration percentage. These results were consistent with other natural compounds like Garcinone E and Luteolin, which inhibited the cell migration in a dose-dependent manner.[31,32] Thus, our findings clearly demonstrate that cytisine significantly inhibit the migratory ability of cancer cells compared to untreated cells, supporting its potential in limiting tumour spread.

CONCLUSION

Molecular docking and dynamic simulation analyses demonstrated a strong binding affinity between cytisine and critical cancer-associated targets, EGFR and TP53, suggesting its potential to modulate key oncogenic pathways. In addition, cytisine exhibited significant anticancer activity against the epithelial carcinoma cell lines. These findings indicate that cytisine may represent a potential anticancer agent; however, additional validation using primary oral carcinoma cell lines and in vivo studies is necessary before establishing its therapeutic relevance.

Conflicts of interest

There are no conflicts of interest.

Funding Statement

Nil.