Breast cancer is one of the most common cancers worldwide and remains a leading cause of cancer-related mortality among women [1]. Despite significant advancements in therapeutic strategies, challenges such as multidrug resistance, tumor progression, and metastasis persist, creating an urgent need for novel approaches to treatment [2].

Two key molecular players implicated in breast cancer are the ATP-binding cassette (ABC) transporters, such as ABCG2, and Protein Tyrosine Phosphatases (PTPs), including Protein Tyrosine Phosphatase 1B (PTP1B). ABCG2 is a critical efflux transporter involved in multidrug resistance. By actively exporting chemotherapeutic agents out of cancer cells, ABCG2 reduces drug efficacy and contributes to treatment failure 3. Targeting ABCG2 has shown potential for overcoming resistance and improving therapeutic responses [4].

Protein Tyrosine Phosphatases (PTPs) are enzymes that regulate various cellular signaling pathways by dephosphorylating tyrosine residues on proteins. Among these, PTP1B has emerged as a significant target due to its role in promoting oncogenic signaling pathways such as HER2 and IGF, which drive tumor growth and metastasis [5,6]. Overexpression of PTP1B is associated with poor prognosis in breast cancer, making it an attractive therapeutic target [7].

Natural compounds derived from medicinal plants offer a promising avenue for breast cancer therapy. These compounds, known for their structural diversity and pharmacological activity, have been shown to interact with critical cancer-related proteins. Medicinal plants such as mandarin peel, date palm seeds, and fenugreek seeds, native to the Chlef region of Algeria, are rich in bioactive compounds such as rutin, quercetin, and naringin, which exhibit anticancer, antioxidant, and anti-inflammatory properties [8, 9]. These phytochemicals have demonstrated potential in modulating ABCG2 activity and inhibiting PTPs, including PTP1B.

Advances in computational biology, particularly molecular docking and pharmacoinformatics, have revolutionized the field of drug discovery. Molecular docking enables the virtual screening of bioactive compounds against specific protein targets, providing insights into their binding affinity and interaction mechanisms [10] By leveraging these tools, this study evaluates the interactions of phytochemicals from mandarin peel, date palm seeds, and fenugreek seeds with ABCG2 and PTPs, including PTP1B, to identify potential inhibitors.

This research highlights the therapeutic potential of plant-derived natural products in addressing breast cancer's critical challenges, emphasizing the integration of traditional medicinal knowledge with modern computational techniques.

Hardware
The computer utilized for the study is equipped with the following specifications: an Intel® Core™ i5-2450M processor running at 2.50 GHz, an HD graphics processing unit (GPU), 8 GB of Random Access Memory (RAM), and operates on the Windows 10 operating system.

Drug-Likeness Prediction

Drug-likeness of the test compound was predicted using three filters provided by the DruLiTo program: Lipinski's rule, Veber rule, and Ghose filter. The input for the analysis was provided in *sdf file format [11].

ADMET Prediction

Pharmacokinetics (ADME) and toxicity predictions for the compound derived from mandarin peel, date palm seeds, and fenugreek seeds officinalis were conducted using the pkCSM web tool (http://biosig.unimelb.edu.au/pkcsm/prediction). The analysis utilized the SMILES format as input [11].

Molecular Docking

The molecular docking study targeted, utilizing high-resolution X-ray crystallographic structures obtained from the Protein Data Bank (PDB), specifically 6-Pyruvoyl Tetrahydrobiopterin Synthase (PDB ID: 3I2B) [12], Protein-Tyrosine Phosphatase Non-Receptor Type 1 (PDB ID: 1XBO) [13], and the Broad Substrate Specificity ATP-Binding Cassette Transporter ABCG2 (PDB ID: 6VXI) [14]. Molecular docking simulations were conducted using Molecular Operating Environment (MOE) software [15], where ligand-receptor interactions were evaluated based on London dG and GBVI/WSA dG scoring functions to estimate binding free energy and solvation effects. The docking protocol was validated through Root Mean Square Deviation (RMSD) analysis, ensuring accuracy, reproducibility, and reliability in predicting ligand binding conformations [16]. These findings provide insights into the molecular recognition mechanisms of potential inhibitors targeting associated pathways, which are implicated in metabolic disorders and cancer [17].

Preparation of Ligand

The preparation of ligands for molecular docking is a crucial step to ensure accurate and reliable predictions of ligand-receptor interactions. In this study, all the ligands selected for their reported anticancer potential, were retrieved from the PubChem databases (https://pubchem.ncbi.nlm.nih.gov) [18] in SDF format and converted into PDB format for docking. The structures were energy minimized using the Merck Molecular Force Field (MMFF94) and *Density Functional Theory (DFT/B3LYP/6-31G) **, ensuring optimal geometry and stability.

Table 1. The major compound from Date palm, seed Fenugreek and mandarin peels

Table 2. The three receptors (proteins) preparation

Drug-likeness prediction

Prediction results of drug-likeness of naringin, rutin from Date palm, seed Fenugreek and mandarin peels are shown in table 3.

ADMET prediction

The prediction results of the ADMET of naringin ,rutin, linoleic acid and lutein are shown in table 3

Table 3. ADMET prediction

Table 5. The energy delta G and inhibition constant of mandarin peels (Kaemrferol, Naringin), date palm seed (lutein) and 6VXI

Table 6. The energy delta G and inhibition constant of date palm seeds (quercetin, ferulic acid), fenugreek (Trigonelline) and 1XBO

Table 7. The energy delta G and inhibition constant of mandarin peels (rutin, lauric acid), fenugreek (7’4- dihydroxyflavon) and 3I2B

Table 8. Interaction details of the three targets (proteins)

Figure 3.The interaction between Rutin,Naringin, Lutein , Tamoxifen (active Principle)  and 6VXI

Figure 3. The interaction between Rutin,Naringin, Linoleic acid, Tamoxifen (active Principle)    and 1XBO

The root-mean-square deviation (RMSD) is a metric used to evaluate the similarity between the predicted and actual conformations of a ligand-receptor complex. A low RMSD value reflects a good fit between the ligand and receptor, indicating the reliability of the docking protocol. An RMSD value below 2 Å is generally considered acceptable in docking studies [11].

In this study, the RMSD value of 1.34 Å, 0.86 Å and 1.28 Å confirms that the docking protocol is valid and that the predicted conformations of the ligand-receptor complexes are accurate [11].

The ΔG score, on the other hand, represents the thermodynamic favorability of the ligand-receptor interaction. A more negative ΔG score corresponds to a stronger interaction and higher binding affinity [11].

The results presented in Tables 4, 5, 6and 7 indicate that Naringin exhibits the strongest inhibitory activity, with a ΔG value of -7.2124 kcal/mol, demonstrating a high binding affinity for the target receptor, PTP1B. In comparison, its interaction with PTPs shows a ΔG value of -5.1495 kcal/mol and -5.7809 kcal/mol with ABGC2. The calculated Ki values about 1000 further support the high binding affinity of rosmarinic acid towards the target receptor.

Inhibitor reference ligand

When we studying the inhibitor reference ligand with in silico programs, the linkages between it and the protein (3I2B) we have got 20% of amino acids (VAL20,GLU62). When we studying the inhibitor reference ligand with in silico programs, the linkages between it and the protein (1XBO) we have got 30% amino acid (ASP  181,ARG221,SER216,CYS215,ARG 254,ARG 24). When we studying the inhibitor reference ligand with in silico programs, the linkages between it and the protein (6VXI) we have got 15% amino acid (PHE  439,ASN436,THR542).

The reference ligand (6VXI)  forms hydrogen bonds with THR 542 (B), PHE 439 (B), and ASN 436 (B) at distances of 2.71 Å, 2.64 Å, and 2.64 Å, respectively, all within the optimal range for stable hydrogen bonding (typically 2.5–3.5 Å). These distances suggest strong and favorable interactions, particularly with THR 542 and PHE 439. The π-π interaction with PHE 439 (B) occurs at 3.79 Å, which is slightly above the ideal range (typically 3.3–3.8 Å), possibly reducing its contribution to binding stability. Overall, the ligand demonstrates well-positioned hydrogen bonds, while the π-π interaction may require structural optimization for better stacking.

The reference ligand (1XBO) forms multiple hydrogen bonds and ionic interactions within the receptor, with distances mostly falling within the optimal range for stability. Strong hydrogen bonds are observed with ASP 181 (A) (2.63 Å), SER 216 (A) (2.86 Å), and ARG 221 (A) (2.63–3.48 Å), suggesting stable interactions. The shortest and likely strongest hydrogen bonds occur with ASP 181 (2.63 Å) and ARG 221 (2.63–2.68 Å). Ionic interactions are also prominent, particularly with ARG 221 (2.63–3.49 Å) and ARG 254 (2.73–3.44 Å), further stabilizing the ligand in the binding site. Some interactions, such as those at 3.21–3.49 Å, are slightly longer, potentially weakening their contribution to binding. Overall, the ligand is well-anchored in the receptor, with strong hydrogen and ionic interactions supporting its binding affinity.

The reference ligand forms (3I2B) two hydrogen bond interactions with the receptor: one with VAL 70 (B) at a distance of 2.84 Å, which falls within the optimal hydrogen bond range (2.5–3.5 Å), suggesting a stable interaction, and another with GLU 62 (B) at 3.14 Å, which is slightly longer but still within an acceptable range for moderate binding stability. The interaction with VAL 70 (B) is likely stronger due to its shorter distance, while the one with GLU 62 (B) may contribute less to overall binding stability. Optimizing the ligand’s positioning could enhance these interactions for better receptor affinity.

The docking analysis (table 8, figure1) revealed that both rutin and naringin exhibited strong interactions with the ASPARAGINE 436 (ASN 436) and GLUTAMINE 398 (GLN 398) residues of ABCG2, forming H-donor interactions with bond distances ranging from 3.10 to 3.37 Å. These interactions suggest a potential inhibitory effect on the ABCG2 transporter, which could reduce drug efflux and enhance intracellular drug retention, thereby overcoming MDR in breast cancer cells. Lutein, however, did not exhibit any significant interactions with ABCG2, indicating a lack of binding affinity for this target.

PTP1B has been recognized as a therapeutic target in breast cancer due to its involvement in tumor growth and metastasis. The docking analysis (table 8, figure2)  Rutin displayed multiple interactions, including H-acceptor bonds with SERINE 216 (SER 216) and ARGININE 221 (ARG 221) at 3.16 Å and 2.77 Å, respectively, along with π-cation and π-H interactions with LYSINE 120 (LYS 120) and PHENYLALANINE 182 (PHE 182), indicating strong binding affinity. Naringin interacted with ASPARTATE 48 (ASP 48) and GLUTAMINE 262 (GLN 262) via H-donor and H-acceptor interactions, while also engaging in π-H stacking with PHE 182 at 4.18 Å. Linoleic acid exhibited multiple ionic interactions with ARG 221 and CYSTEINE 215 (CYS 215), with distances ranging from 2.88 to 3.91 Å, suggesting a distinct inhibitory mechanism compared to flavonoids. These interactions highlight the potential of these compounds to modulate PTP1B activity, thereby interfering with breast cancer-associated signaling pathways.

In addition to PTP1B, other protein tyrosine phosphatases (PTPs) contribute to oncogenic signaling in breast cancer.  Rutin and naringin interacted with THREONINE 67 (THR 67) and VALINE 70 (VAL 70) via H-donor interactions (3.21–3.49 Å), while linoleic acid established both H-acceptor and ionic interactions with ARGININE 9 (ARG 9) at distances of 2.95 to 3.89 Å (table 8, figure2) . These findings suggest that flavonoids and fatty acids might regulate PTP signaling networks in breast cancer cells, offering potential therapeutic benefits.

Tamoxifen (active Principle) of (Pharmaceutical commercial drug )

When studying the active ingredient of tamoxifen with in silico programs, the linkages between it and the protein (6VXI) are the following amino-acids we have (GLN398) with a ΔG value of -5.51 kcal/mol. When studying the active ingredient of Tamoxifen with in silico programs, the linkages between it and the protein (IXBO) are the following amino-acids ( ASP48, ARG24) with a ΔG value of -5.19 kcal/mol,. When studying the active ingredient of Tamoxifen with in silico programs, the linkages between it and the protein (3I2B) are no amino-acids with a ΔG value of -4.93 kcal/mol,

The study results indicate that the compounds derived from Date palm, seed Fenugreek and mandarin peels have potential as drug candidates due to their favorable pharmacokinetic properties and strong binding affinity towards the receptors 3I2B, 6VXI, and 1XBO, with a particular emphasis on PTP1B (1XBO). Molecular docking analysis identifies Rutin and Naringin as the most potent inhibitor, exhibiting the highest affinity for its target receptor. Additionally, adherence to Lipinski's and Weber's rules further supports their potential for drug development.

1. (2021). Global Breast Cancer Statistics. Retrieved from https://www.who.int/.
2. Siegel, R. L., Miller, K. D., & Wagle, N. S. (2023). Cancer statistics, 2023. CA: A Cancer Journal for Clinicians, 73(1), 17–48.
3. Gottesman, M. M., & Ling, V. (2006). The molecular basis of multidrug resistance in cancer: The ABC transporters. Nature Reviews Cancer, 6(2), 93–106.
4. Polgar, O., Bates, S. E., & Robey, R. W. (2020). ABCG2: Structure, function, and role in drug response. Trends in Pharmacological Sciences, 41(3), 178–190.
5. Tonks, N. K. (2006). Protein tyrosine phosphatases: From genes, to function, to disease. Nature Reviews Molecular Cell Biology, 7(11), 833–846.
6. 6Julien, S. G., Dube, N., Hardy, S., & Tremblay, M. L. (2007). Inside the human cancer tyrosine phosphatome. Nature Reviews Cancer, 7(1), 35–49.
7. 7Zhang, Z. Y., & Van Goor, F. (2018). Targeting PTP1B in cancer. Nature Reviews Drug Discovery, 17(3), 183–196.
8. 8Harborne, J. B., Baxter, H., & Moss, G. P. (1999). Phytochemical Dictionary: A Handbook of Bioactive Compounds from Plants. Taylor & Francis.
9. 9Ayoub, F., Fawaz, M., & Baydoun, S. (2021). Phytochemical constituents of Algerian medicinal plants and their bioactivity. Journal of Ethnopharmacology, 268,
10. 10Morris, G. M., Huey, R., & Olson, A. J. (2009). Using AutoDock for ligand-receptor docking. Current Protocols in Bioinformatics, 24(1), 8-14.
11. Hammoudi, Mounir, Khaled Otmanine, and Brahim Feraoun. "In Silico Analysis and ADMET Prediction of Cinnamon Compounds on PPARα/γ Receptor for Searching AntiDiabetic Drug Candidates." Journal of Jilin University (Engineering and Technology Edition), 2023, doi:10.1671-5497.
12. Erlandsen, H., Bjørgo, E., Flatmark, T., & Stevens, R. C. (2000). Crystal structure and properties of 6-pyruvoyl tetrahydropterin synthase. Journal of Molecular Biology, 297(2), 299–310. https://doi.org/10.1006/jmbi.2000.3569
13. Salmeen, A., Andersen, J. N., Myers, M. P., Tonks, N. K., & Barford, D. (2003). Molecular basis for the dephosphorylation of the activation segment of the insulin receptor by protein tyrosine phosphatase 1B. Molecular Cell, 11(6), 1551–1562. https://doi.org/10.1016/S1097-2765(03)00220-2
14. Manolaridis, I., Jackson, S. M., Taylor, N. M. I., Kowal, J., Stahlberg, H., & Locher, K. P. (2018). Cryo-EM structures of a human ABCG2 mutant trapped in ATP-bound and substrate-bound states. Nature, 563(7731), 426–430. https://doi.org/10.1038/s41586-018-0680-3
15. Chemical Computing Group. (2021). Molecular Operating Environment (MOE), Version 2021. Montreal, Canada: CCG Inc.
16. Trott, O., & Olson, A. J. (2010). AutoDock Vina: Improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. Journal of Computational Chemistry, 31(2), 455–461. https://doi.org/10.1002/jcc.21334
17. Berger, J., & Moller, D. E. (2002). The mechanisms of action of PPARs. Annual Review of Medicine, 53(1), 409–435. https://doi.org/10.1146/annurev.med.53.082901.104018
18. Vijayakumar, K., et al. "Anti-diabetic and Hypolipidemic Effects of Cinnamon Cassia Bark Extracts: An In Vitro, In Vivo, and In Silico Approach." Journal of Food Science and Technology, 2020, doi:10.1080/10408398.2020.1763904.