Translate this page into:
Novel isoxazole-containing thiazolidinone: Synthesis, DFT analysis, and pharmacological evaluation
*Corresponding author: E-mail address: s.alosaimi@tu.edu.sa (S. Alotaibi)
-
Received: ,
Accepted: ,
Abstract
The monoterpene skeleton of (R)-Carvone was selected as a support to access the new heterocyclic isoxazole ring combined with thiazolidinone via the Huisgen cycloaddition reaction. A good yield was achieved for the synthesized product, which was characterized by high-resolution mass spectrometry (HRMS) and NMR (1H & 13C) analysis. The chemoselectivity of the 1,3-dipolar cycloaddition reaction indicated that the dipolarophilic site is more reactive than the double bond. Density functional theory (DFT) was utilized to determine the peri- and regio-selectivity of the [3+2] cycloaddition reaction of the compounds. The theoretical outcomes closely align with the experimental results. Additionally, the electronic interaction properties of the synthesized compounds were examined through the study of frontier orbitals highest occupied molecular orbital (HOMO), lowest unoccupied molecular orbital (LUMO) and independent gradient model (IGM) analysis. Furthermore, the effect of various solvents on the λmax of the two regioisomers 6 and 6T was investigated. Using network pharmacology, critical genes and pathways associated with breast cancer were identified, highlighting epidermal growth factor receptor (EGFR) and poly (adp-ribose) polymerase 1 (PARP1) as potential therapeutic targets for the designed compounds. Molecular docking studies were then conducted with these top targets, EGFR and PARP1, to assess the binding affinity of the ligands, demonstrating strong interactions between compound 6 and these proteins.
Keywords
DFT
Isoxazole
Molecular docking
Network pharmacology
(R)-carvone
Regioselectivity
1. Introduction
Cancer is characterized by uncontrolled cell proliferation and its potential to metastasize to distant parts of the body, making it the second leading cause of mortality worldwide [1,2]. Breast cancer, in particular, can grow rapidly and aggressively, with cancerous cells originating in one location, such as the blood, and spreading to other tissues [3]. In liver cancer, hepatocellular carcinoma (HCC) accounts for over 70% of primary liver cancer cases globally, based on statistical data [4,5].
Monoterpenes, natural compounds widely present in essential oils (up to 90% of the composition), have long been utilized in human care products such as fragrances, flavoring agents, and cosmetics [6,7]. These compounds exhibit diverse therapeutic properties, including anti-inflammatory [8], antinociceptive [9], and vasorelaxant effects [10]. In medicinal chemistry, five-membered heterocycles are highly valued for their biological activity and serve as promising scaffolds for drug development [11].
One notable heterocycle is the isoxazole ring, which has attracted considerable attention owing to its five-membered structure and the presence of heteroatoms (oxygen and nitrogen) that can form hydrogen bonds with biological targets [12]. Isoxazole derivatives are recognized for their diverse biological activities, including anticancer properties [13]. Moreover, these derivatives find applications in organic chemistry as intermediates and in biology for therapeutic purposes [14]. For instance, Ait Itto et al. synthesized a series of isoxazoline derivatives exhibiting potent antiproliferative effects against cancer cells. Their apoptotic studies revealed that these compounds induce cell death through caspase-3/7-mediated apoptosis [15]. Isoxazole-based medications, including Risperidone, Leflunomide, Cycloserine, Sulfamethoxazole (an antibacterial), and Acivicin (an antitumor agent), are currently in clinical use, underscoring the pharmacological significance of this scaffold [16]. In drug development, the anticancer agent Luminespib (resorcinylic isoxazole amide NVP-AUY922) is currently undergoing 28 phase I/II clinical trials. Preclinical studies indicate that isoxazoles display antitumor activity primarily through the inhibition of heat shock protein 90 (HSP90) [17]. Several synthetic methods for isoxazoles are documented, with the most prevalent The [3+2] cycloaddition reactions of nitrile oxides with alkynes or alkenes, along with intramolecular cyclisation methods [18].
The epidermal growth factor receptor (EGFR) is a cell-surface transmembrane protein vital for regulating cell proliferation and differentiation. When epidermal growth factor (EGF) or other ligands bind to EGFR, it activates signaling pathways that promote cell division and growth. To disrupt these pathways and inhibit tumor cell proliferation, EGFR inhibitors have been developed as targeted therapies [19]. Monoclonal antibodies function by binding to the extracellular region of EGFR, blocking ligand interaction and subsequent signaling. In contrast, small molecule inhibitors act within the intracellular tyrosine kinase domain, directly suppressing enzymatic activity and downstream signaling cascades [20]. Due to their improved efficacy and tolerability compared to traditional chemotherapy, anti-EGFR therapies are now recommended as first-line treatments for advanced cancers with EGFR mutations, as outlined in international clinical guidelines [21].
Although significant research has been conducted in anticancer drug discovery, the potential of isoxazole-thiazolidinone derivatives derived from (R)-carvone remains largely underexplored. Computational approaches have proven indispensable for understanding protein folding and unfolding, as well as predicting protein-ligand and protein-protein interactions [22]. These methods provide crucial insights for designing rational drug candidates by modelling diverse protein conformations, including both active and inactive states. Integrating computational predictions with experimental findings enhances the accuracy and depth of understanding protein behavior, enabling the refinement of drug design strategies and bridging the gap between theory and practice.
However, despite the significant activity of reported isoxazole and thiazolidinone analogs [13-17], the literature reveals a structural gap concerning the conjugation of such a versatile heterocyclic core with a chiral, low-toxicity monoterpene scaffold, such as (R)-Carvone. The studies presented above [23,24] have primarily focused on classical aliphatic or aromatic substitutions, overlooking the exploration of hybrids incorporating a chiral monoterpene motif a domain largely underexplored in chemotherapy. Our approach aims to address this void by combining the hydrogen-bonding properties of the isoxazole-thiazolidinone moiety with the three-dimensional structure and associated low toxicity of (R)-Carvone. This novel Carvone-isoxazole-thiazolidinone scaffold is structurally distinct from the cited analogs, offering a unique opportunity to enhance the selectivity and absorption, distribution, metabolism, excretion, and toxicity (ADMET) profile of future anticancer drug candidates.
Building on prior studies that identified heterocyclic ring-based inhibitors derived from (R)-carvone, this work takes a novel, multidisciplinary computational approach to evaluate the interaction between (R)-carvone-isoxazole derivatives and specific apoptotic protein targets. The study explores the therapeutic potential of these compounds, drawing from the well-documented role of isoxazole moieties as critical pharmacophores in anticancer agents. As part of our ongoing efforts to discover new anticancer compounds [23,24], we have successfully synthesized an (R)-carvone-thiazolidinone-isoxazole hybrid, marking an important step in evaluating these derivatives as potential therapeutic candidates.
2. Materials and Methods
NMR analyses were conducted using a Bruker avance III spectrometer (operating at 300 and 75 MHz). using CDCl3 as the solvent. Molecular weight was determined using a Q-TOF micro mass spectrometer. It is important to mention that the intermediate products (2, 3, and 4) have been previously described in our earlier work [25].
2.1. The preparation of isoxazole 6
At 0 °C, dipolarophile 4 (0.5 g, 1.66 mmol) and dipole 5 (0.8 mmol) were dissolved in 25 mL of dichloromethane. A 25 mL solution of sodium hypochlorite (NaOCl, 1.66 mmol) was then added dropwise to the reaction mixture while maintaining the same temperature. The mixture was stirred continuously for one hour. The product 6 was purified using column chromatography on silica gel, employing a mixture of ethyl acetate and hexane (10/90) as the eluent.
2-(2-Methyl-5-(propen-2yl)cyclohex-2-en-ylidene)hydrazono-3-((3-phenylisoxazol-5-yl) methyl)thiazolidinone 6. Yield 78%, physical aspect is white solid, mp= 161-163 °C, Rf: 0.25, (Hexane/Acetate, 1/9), For HRMS: value found 421.1632 [M+H]+, calculated value 421.1646 (Figure 1S). 1H-NMR: 1.72 (3H, s, CH3 ); 1.93 (3H, s, CH3); 2.14 (1H, m, CH); 2.14-3.33 (4H, m, 2×CH2); 3.84 (2H, s, CH2); 4.75 (2H, s, CH2); 5.17 (2H, s, =CH2 ), 6.23 (1H, m, =CH), 6.58 (1H, s, =CH), 7.44-7.45 (3H, m, HCAr ), 7.75-7.79 (2H, m, HCAr) (Figure 2S). 13C-NMR : 17.68 (CH3); 20.64 (CH3); 27.98 (CH2); 32.31 (CH2); 31.62 (CH2); 38.16 (CH2); 41,08 (CH); 101.60 (=CH); 109.99 (=CH2); 126.81 (HCAr); 128.84 (HCAr); 130.03 (HCAr); 136.05 (=CH; 147.56 (C); 147.89 (CAr ); 151.72 (C); 158.60 (C); 162.48 (C=N); 202.76 (C=O) (Figure 3S).
2.2. Computational details
The computational analysis was performed using GAUSSIAN 09W software with the DFT method B3LYP/6-31G’(d, p). The presence of only positive vibrational frequencies confirmed that the optimized structure corresponds to a minimum energy configuration. The chemical hardness (η) and chemical potential (μ) were determined from the energies of the HOMO (E_H) and LUMO (E_L) orbitals using the relations: η = (E_L − E_H) and μ = (E_H + E_L)/2.The global nucleophilicity (N) and electrophilicity (ω) indices were then obtained as N = E_H − E_H and ω = μ2/2η, respectively.
The Mulliken atomic spin densities were used to evaluate the Parr function, while the global electron density transfer (GEDT) was estimated by summing the natural atomic charges (q) over the nucleophilic framework atoms at the transition states: GEDT = Σq_A. The physical descriptor δ_g was employed to describe the Independent Gradient Model (IGM) analysis, defined as δ_g = |∇ρ^IGM| - |∇ρ|. Here, the electron density (ρ) acts as a local and measurable scalar field, where |∇ρ^IGM| represents the theoretical maximum gradient and |∇ρ| corresponds to the gradient in a non-interacting system.
2.3. Network pharmacology
A systematic approach was employed to identify potential target genes for the Thiazolidinone-isoxazole derivatives (6 & 6T) in relation to breast cancer. Swiss target prediction was initially utilized to discover possible human targets for these compounds via SMILES notation [26,27], followed by gene analysis. The DisGeNET database was then employed to filter genes associated with breast cancer [28,29], with overlaps confirmed using Venn diagrams [30].
Functional enrichment analysis was conducted using the DAVID bioinformatics tool to explore the biological significance of the shared target genes. Gene Ontology (GO) categorized these genes into Biological processes (BP), Molecular functions (MF), and Cellular components (CC), while kyoto encyclopedia of genes and genomes (KEGG) pathway analysis identified key signaling pathways related to antibacterial activity (p ≤ 0.05) [31-33]. A compound–target interaction network was built in Cytoscape (v3.10.0), where nodes represented genes and edges denoted interactions. Genes with the highest connectivity were defined as hub genes, and the top ten targets were selected for further analysis. To validate these findings, a protein–protein interaction (PPI) network was generated using the STRING database, and hub genes were ranked using CytoHubba based on topological parameters such as degree and betweenness centrality, highlighting key regulators involved in antibacterial mechanisms [34-36].
2.4. Physicochemical properties
The pharmacological efficacy of a drug is closely linked to its physicochemical properties, which influence its pharmacokinetics. Key properties such as Log P, molecular weight, hydrogen bond donors/acceptors, and topological polar surface area (TPSA) are essential for evaluating a compound’s absorption, distribution, and bioavailability. These properties can be assessed using tools like Molinspiration software. According to Lipinski’s Rule of 5, compounds with Log P ≤ 5, molecular weight ≤ 500, ≤ 10 hydrogen bond acceptors, and ≤ 5 hydrogen bond donors typically demonstrate good bioavailability. Deviations from these parameters may indicate potential absorption or therapeutic issues [37-39].
2.5. Bioactivity
Bioactivity scores for the designed compounds were calculated using Molinspiration software, an essential tool in drug development. These scores forecast the compounds’ potential efficacy against specific targets [39].
2.6. Toxicity
Advancements in computational methods have greatly enhanced toxicity prediction, facilitating in silico safety assessments of chemical compounds. The ProTox-II server allows for effective prediction and analysis of toxicity profiles for newly designed compounds, promoting safer drug development and minimizing the need for extensive in vitro or in vivo testing [40-43].
2.7. Molecular docking
Molecular docking was conducted using AutoDock 4.2 [44]. The ligand was imported into UCSF Chimera through its SMILES notation, converted into a 3D structure, and energy-minimized to generate a PDB file [45]. The top two targets identified from network pharmacology, EGFR (PDB ID: 4HJO) and poly (ADP-ribose) polymerase 1 (PARP1) (PDB ID: 5HA9), were retrieved from the RCSB Protein Data Bank. The protein preparation involved removing non-essential residues and water molecules, adding polar hydrogens and Kollman charges, and converting the structure to PDBQT format. A grid box was generated, and docking was performed using the Lamarckian genetic algorithm. Results were analyzed using BIOVIA Discovery Studio for the visualization of ligand-protein interactions in 2D and 3D [46-51].
3. Results and Discussion
3.1. Chemistry
A useful method for synthesizing isoxazole derivatives substituted with an aryl group involves the cyclisation of thiosemicarbazone-carvone 2 with ethyl 2-bromoacetate in a basic medium (sodium acetate), resulting in an 86% yield of thiazolidinone 3. The next step in our synthesis is the N-alkylation of the thiazolidinone. Using propargyl bromide and K2CO3, the reaction was conducted at room temperature, yielding compound 4 with an 89% yield (Scheme 1).
- Operating conditions and reagents: (i) Thiosemicarbazide, H2SO4, reflux, EtOH, (ii) Ethyl bromoacetate, AcONa, EtOH, reflux. (iii) Propargyl bromide, K2CO3, RT, acetone.
The compound 4 was treated with benzaldehyde oxime in the presence of a sodium hypochlorite (NaOCl) solution. It is important to note that dipolarophile 4 contains several reactive sites, including a triple bond, two C=N bonds, and two C=C double bonds. Consequently, various heterocyclic systems, such as isoxazoline, isoxazole, and/or oxadiazole, can be synthesized through cycloaddition reactions. This reaction may also result in a regioselective product (Scheme 2).
![The theoretical possibility of obtaining several five-membered rings from of this [3+2] cycloaddition reaction of compound 4 with arylonitrile 5.](/content/184/2026/19/4/img/AJC-19-5552025-g3.png)
- The theoretical possibility of obtaining several five-membered rings from of this [3+2] cycloaddition reaction of compound 4 with arylonitrile 5.
The reaction mixture was extracted using dichloromethane, dried over anhydrous Na2SO4, and purified via silica gel chromatography. The isoxazole derivative 6 was obtained with a yield of 78%, demonstrating complete peri- and regioselectivity (Scheme 3).
- Synthesis of novel (R)-carvone-isoxazole.
The isoxazoline derivative 6 was characterized by HRMS and NMR, with the HRMS spectrum showing an ion at m/z = 421.1632, matching the molecular formula C23H24N4O2S. The NMR spectrum comparison between compounds 4 and 6 revealed the disappearance of the acetylenic moiety (δ 1H 2.18 ppm; δ 13C 71.53 and 76.97 ppm). However, signals for the C=CH (endocyclic double bond: δ 1H 6.23 ppm; δ 13C 130.56 ppm) and C=CH2 (exocyclic double bond: δ 1H 4.75 ppm; δ 13C 109.99 ppm) were retained, along with signals for C=N (C=N-Carvone: δ 13C 162.48 ppm) and C=N (C=N-thiazolidinone: δ 13C 147.89 ppm). The NMR spectra (1H and 13C) confirm that dipole 5 attacked the triple bond in a periselective and regioselective manner. The CH2 groups of the thiazolidinone core and the N-CH2 group appear at (δ 1H: 3.82 ppm; δ 13C: 31.62 ppm) and (δ 1H: 5.17 ppm; δ 13C: 38.16 ppm), respectively. In contrast, the methyl (CH3) groups from (R)-carvone are observed at (δ 1H: 1.72 ppm; δ 13C: 17.68 ppm) and (δ 1H: 1.93 ppm; δ 13C: 20.64 ppm). The peak for the 1,2,3-triazole core (HC=C) is found at (δ 1H: 6.58 ppm; δ 13C: 101.60 ppm). The 1H-NMR spectrum correlates each peak with its corresponding dipolarophile site (Figure 1S).
3.2. Mechanistic study
The problem that arises in the 1,3-dipolar cycloaddition reaction we are investigating is defined at the level of chemo-selectivity and regioselectivity. This means that the two atoms of the dipoles (oxygen and carbon) can add to the double bonds C=C, the triple bond, and/or the C=N bond of dipolarophile 4. The calculation of all theoretical possibilities yielded eight products, labelled PA1, PA2, PB1, PB2, PC1, PC2, PD1, and PD2 (this refers to the numbering of all chemical possibilities) (Scheme 4). Given the number of transition states to be calculated for this work, we have eight transition states abbreviated as TSA1, TSA2, TSB1, TSB2, TSC1, TSC2, TSD1, and TSD2. Our objective in this mechanistic study using DFT is to investigate the dipole’s approach to the dipolarophile, determine the energy barrier of each transition state, trace the most favorable reaction pathway, and ultimately describe the electronic density [32].
- Proposed problem regarding the reactivity of oxime R1 with dipolarophile P1.
The B3LYP/6-31G’(d, p) level of DFT was used to optimize the geometric structures of compound (PA1) as showed in Figure 1.
- Optimized structures of the compounds (PA1) at B3LYP/6-31G’(d, p) level.
The measurement of the global electron density transfer (GEDT) value is used to assess global reactivity indexes, which can explain the reaction mechanism [32]. According to Table 1, the comparison between the calculated values of electronic chemical potential (µ) for compounds R1 (-5.263 eV) and P1 (-3.558 eV) indicates a transfer of dipolarophilic charge to the dipole. Furthermore, to substantiate these results, the calculation of η for both the dipole and dipolarophile revealed that compound R1 (η=2.595 eV) possesses a significant capacity to donate electrons compared to P1 (4.366 eV).
| Physical parameters | HOMO | LUMO | μ | η | ω | N |
|---|---|---|---|---|---|---|
| P1 | -5.741 | -1.375 | -3.558 | 4.366 | 1.450 | 3.375 |
| R1 | -6.560 | -3.965 | -5.263 | 2.595 | 5.335 | 2.556 |
Through the literature, we have concluded that type 1,3-dipolar cycloaddition reactions possess a regioisomeric channel, wherein bond formation occurs when the most electrophilic and nucleophilic centers of the reagents meet [52,53]. It would be desirable to have local reactivity indexes capable of describing these relevant centers in organic molecules [54]. 1,3-dipolar cycloaddition reactions have shown that examining local electrophilicity (ωk) and nucleophilicity (Nk) at both the electrophilic reagent and the nucleophilic site (Parr functions), respectively, enables clarification of the experimentally observed regioselectivity. Table 2 summarizes the values of the nucleophilic Parr functions, electrophilic parameters, as well as the local nucleophilicity and local electrophilicity at R1 and P1.
| Physical parameters | Number of atoms | Pk+ | PK- | ωk | Nk |
|---|---|---|---|---|---|
| R1 | O1 | 0,0848 | 0,4511 | 0,4526 | 1,5225 |
| N2 | 0,1770 | 0,0626 | 0,9441 | 0,2111 | |
| C3 | -0,0230 | 0,0232 | -0,1229 | 0,0782 | |
| P1 | C20 | -0,0010 | 0,0038 | -0,0014 | 0,0128 |
| C21 | 0,0003 | 0,0026 | 0,0004 | 0,0066 |
The analysis of the data regarding local nucleophilicity Nk at R1 indicates that the oxygen atom O1 is the most nucleophilic atom involved in this 1,3-dipolar cycloaddition reaction, with a value of NO1=1.52 eV. Conversely, the analysis of the data concerning local electrophilicity at P1 reveals that the C21 carbon is the most nucleophilic center, with a value of ωC21=0.0004 eV. Consequently, the first sigma bond to form is the O1-C21 bond, accompanied by a significant percentage of electrophilic activation at the N2 nitrogen, ωN2=0.91 eV. For the C20 carbon, there is the largest nucleophilic activation (NC20=1.52 eV at P1) (Figure 2 and Table 2).
- The calculation is made via the Parr function (ev) and for the red represente local electrophilicites (ωk) and the blue represente local nucleophilicites (Nk).
In another part, the various products have associated activation barriers: PA1, PA2, PB1, PB2, PC1, PC2, PD1, and PD2, which are equal to 12.15, 19.20, 33.27, 57.61, 59.13, 14.41, and 23.31 (Kcal. mol-1) respectively, as presented in Figure 3. According to the figure above, the PA1 compound exhibits the lowest activation energy among the other compounds. Consequently, the product PA1 is deemed the most favored product kinetically. Furthermore, in terms of thermodynamic stability, PA1 is the most stable compound. The stability difference between the most stable product PA1 (-79.61 Kcal. mol-1) and the least stable product PB2 (-11.61 Kcal. mol-1) is 68 Kcal. mol-1. This theoretical analysis identifies a single product (PA1) with low activation energy from eight possibilities proposed within the scope of this work, which aligns well with the experimental results (Figure 3). Figure 4 illustrates the optimized geometries of the transition states (TSs), corresponding to the eight reaction channels presented in Scheme 3. Regarding product PA1, the lengths of the new sigma bonds formed, such as C-C and O-C, at the transition state (TSA1) are 2.22 Å and 2.39 Å respectively with an imaginary frequency of -1901.800167 harte. It is important to note that this represents the lowest activation barrier value.
- Calculation of energy profiles (Kcal.mol-1) of all possibilities between the dipole with the dipolarophile.
- Optimized geometries of the TSs via B3LYP/6-31G’(d, p) were obtained from the 1,3-dipolar cycloaddition between oxime and dipolarophile. Furthermore, the distances between atoms have been calculated in angstroms (Å).
3.3. Independent gradient model
Drawing on a study by Yang et al. that examines non-covalent interactions and their visualization, this research has become critical in understanding the underlying mechanisms of both attractive and repulsive interactions through the reduced density gradient (RDG) isosurface representation. In the current work, applying this approach yielded a value of 0.02 a.u for the IGM gradient isosurface, as illustrated in Figure 5. This isosurface also encompasses electron density values corresponding to the formation of new σ bonds during the [3+2] cycloaddition reaction of compound 4. Figure 6 displays a comprehensive isosurface illustrating the interaction region among all potential transition state fragment interactions. This analysis uncovers significant steric hindrance in all transition states except TSA1 and TSD1, with TSA1 demonstrating greater stabilization compared to TSD1, thus suggesting a preferential pathway via isoxazoline formation, which aligns with experimental observations.
- Intermolecular interaction during the TSs with the help of analysis of Independent gradient model (IGM).
- Effect of solvent on the UV spectra of structures 6 and 6T.
Concerning the competition between the formation of two new σ bonds, it is established that the electron density of the C-C bond is considerably higher than that of the O-C bond. As a result, during bond formation, the C-C σ bond is formed first, followed by the O-C σ bond. Theoretical values derived from IGM analysis indicate electron density values for the C-C and O-C bonds of 0.4618 and 0.2866, respectively. These findings suggest that the formation of the C-C bond occurs more swiftly than that of the O-C bond. Therefore, one can conclude that the formation of the two new σ bonds occurs asynchronously, in accordance with both the kinetic and thermodynamic analysis of compound 4.
3.4. Effect of solvent on the UV spectra
The data presented in Table 3 (and Figure 6) provide insights into the UV absorption spectra of molecules 6 and 6T across different solvents and in the gas phase, focusing on maximum absorption wavelength (λmax), excitation energy (Eex), oscillator strength (ƒ), and the HOMO→LUMO transition. Here are a scientific analysis and comparison of the results:
| Solvent | Structure | λmax | ƒ | Eex | Assignment |
|---|---|---|---|---|---|
| Chloroform | 6 | 349 | 0.0077 | 3.56 | HOMO→LUMO (99%) |
| 6T | 345 | 0.0022 | 3.59 | HOMO→LUMO (96%) | |
| DMSO | 6 | 325 | 0.0027 | 3.81 | HOMO→LUMO (97%) |
| 6T | 323 | 0.0015 | 3.83 | HOMO→LUMO (98%) | |
| Water | 6 | 324 | 0.0023 | 3.81 | HOMO→LUMO (97%) |
| 6T | 323 | 0.0013 | 3.83 | HOMO→LUMO (98%) | |
| Dichloromethane | 6 | 325 | 0.0065 | 3.80 | HOMO→LUMO (96%) |
| 6T | 323 | 0.0019 | 3.82 | HOMO→LUMO (98%) | |
| Gas | 6 | 329 | 0.0071 | 3.76 | HOMO→LUMO (93%) |
| 6T | 327 | 0.0012 | 3.79 | HOMO→LUMO (98%) |
Molecule 6 consistently exhibits a slightly higher λmax than molecule 6T across all solvents, indicating that 6 absorbs light at longer wavelengths than 6T. In chloroform, the λmax of molecule 6 is 349 nm, whereas for 6T, it is 345 nm. This trend continues in polar solvents (dimethyl sulfoxide (DMSO) and water), where λmax for 6 is 324-325 nm compared to 323 nm for 6T. In the gas phase, λmax values stand at 329 nm for 6 and 327 nm for 6T. The slight red shift (longer wavelength) for molecule 6 suggests that it has a lower energy transition compared to 6T, which may result from subtle structural differences between the two molecules, leading to a stronger stabilization of the excited state in molecule 6. As anticipated, the Eex values are inversely related to λmax. with molecule 6 exhibiting lower excitation energies than molecule 6T. For instance, in chloroform, Eex is 3.56 eV for molecule 6 and 3.59 eV for molecule 6T. In DMSO and water, Eex for molecule 6 is 3.81 eV, whereas for 6T, it is 3.83 eV. These differences indicate that molecule 6 requires slightly less energy for electronic excitation, reinforcing the observation that it absorbs at longer wavelengths. Molecule 6 displays significantly higher oscillator strengths (ƒ) compared to molecule 6T across all environments. In chloroform, for example, ƒ is 0.0077 for molecule 6 versus 0.0022 for molecule 6T. This pattern persists across solvents and in the gas phase. Higher ƒ values for molecule 6 indicate more intense electronic transitions, suggesting that it may engage more effectively with light, which could be advantageous for applications requiring strong optical absorption.
3.5. Comparison between solvent and gas phase
Both molecules exhibit solvent-dependent shifts in λmax and Eex. Polar solvents (e.g., DMSO, water) lead to slightly blue-shifted absorption (lower λmax and higher Eex) compared to nonpolar solvents (e.g., chloroform, dichloromethane). In the gas phase, both molecules show higher λmax values (329 nm for 6 and 327 nm for 6T) than those in polar solvents, indicating that the absence of solvent stabilization slightly lowers the excitation energy. The HOMO→LUMO transition predominates in both molecules, with contributions ranging from 93% to 99%. Molecule 6 shows a slightly lower percentage in the gas phase (93%), whilst molecule 6T maintains a high contribution (98%) across all environments. This suggests a consistently localized electronic transition for molecule 6T, whereas molecule 6 exhibits slightly more delocalization in the gas phase. Molecule 6 demonstrates stronger electronic transitions, longer wavelength absorption, and lower excitation energies compared to molecule 6T, indicating a slight electronic advantage that may render it more suitable for applications involving UV absorption. Both molecules are sensitive to the polarity of the solvent, with polar solvents inducing blue shifts in λmax due to stronger solvent-molecule interactions. The higher oscillator strength of molecule 6 emphasizes its superior light absorption properties, potentially making it more effective in optoelectronic applications.
Key electronic and quantum properties, including the energy gap (HLG), chemical hardness (η), and chemical potential (μ), were calculated using Equations 1-4.
The data in Table 4 and Figure 7 examines the electronic properties of molecules 6 and 6T, specifically the HOMO and LUMO energies, HOMO-LUMO gap (HLG), chemical potential (μ), and chemical hardness (η), across various solvents and in the gas phase. The HOMO (highest occupied molecular orbital) and LUMO (lowest unoccupied molecular orbital) energies indicate the electron-donating and -accepting abilities of the molecules, respectively. Molecule 6 generally has a slightly higher HOMO energy than 6T across all environments. For example, in chloroform, the HOMO of 6 is -5.69 eV compared to -5.84 eV for 6T. Similarly, in DMSO, the HOMO values are -5.97 eV and -5.98 eV for 6 and 6T, respectively. LUMO energies are also higher for molecule 6 compared to 6T. In chloroform, the LUMO of 6 is -1.38 eV, while for 6T it is -1.48 eV. This trend persists across all solvents. Molecule 6T consistently shows slightly higher HLG values compared to molecule 6 across all solvents, indicating a marginally more stable electronic structure. In chloroform, the HLG for 6T is 4.36 eV, while for 6 it is 4.31 eV. In DMSO, water, and dichloromethane, the HLG for 6T is 4.5 eV, compared to 4.49 eV for 6. In the gas phase, these values are 4.51 eV for 6T and 4.48 eV for 6. Molecule 6 exhibits slightly more negative chemical potential values compared to 6T in all solvents, indicating a stronger tendency to attract electrons. In chloroform, the μ of 6 is -3.53 eV, while for 6T it is -3.66 eV. In DMSO and water, the values for 6 and 6T are approximately -3.72 eV and -3.73 eV, respectively. Chemical hardness values for both molecules are similar across all solvents, with 6T exhibiting marginally higher η values. This suggests that 6T is slightly more resistant to electronic deformation. In DMSO and water, η is 2.25 eV for 6T and 2.24 eV for 6. In the gas phase, η is 2.25 eV for 6T and 2.24 eV for 6. Polar solvents such as DMSO and water stabilize both molecules, leading to higher HLG and η values compared to nonpolar solvents like chloroform and dichloromethane. In the gas phase, HLG values for both molecules are slightly higher compared to solvent environments, indicating reduced stabilization of the electronic structure in the absence of solvent effects. Molecule 6 exhibits higher HOMO and LUMO energies, slightly lower HLG values, and more negative chemical potential, indicating a stronger electron-attracting ability but less electronic stability compared to molecule 6T. Molecule 6T shows slightly higher HLG and chemical hardness values, suggesting it has a more stable electronic structure and is less reactive to electronic perturbation. The differences in solvent effects between the two molecules are minor, with polar solvents inducing similar stabilizing trends. These findings highlight subtle electronic distinctions between the molecules, with molecule 6 being potentially more reactive and molecule 6T being more stable.
| Solvent | Structure | LUMO | HOMO | HLG | η | μ | ESolv |
|---|---|---|---|---|---|---|---|
| Chloroform | 6T | -1.48 | -5.84 | 4.36 | 2.18 | -3.66 | -17.40 |
| 6 | -1.38 | -5.69 | 4.31 | 2.15 | -3.53 | -17.55 | |
| DMSO | 6T | -1.48 | -5.98 | 4.5 | 2.25 | -3.73 | -17.12 |
| 6 | -1.48 | -5.97 | 4.49 | 2.24 | -3.72 | -17.09 | |
| Water | 6T | -1.48 | -5.98 | 4.5 | 2.25 | -3.73 | -17.14 |
| 6 | -1.48 | -5.97 | 4.49 | 2.24 | -3.72 | -17.10 | |
| Dichloromethane | 6T | -1.47 | -5.97 | 4.5 | 2.25 | -3.72 | -17.15 |
| 6 | -1.48 | -5.96 | 4.48 | 2.24 | -3.72 | -17.11 | |
| Gas | 6T | -1.38 | -5.89 | 4.51 | 2.25 | -3.63 | ---- |
| 6 | -1.47 | -5.95 | 4.48 | 2.24 | -3.71 | ---- |
- Changes in the solvation-free energy (Esolv) in different solvents for molecules 6 and 6T.
3.6. In-silico study
3.6.1. Network pharmacology
The potential target genes for the synthesized Thiazolidinone-isoxazole derivatives (6 & 6T), comprising 140 genes, were retrieved from the Swiss target database. Additionally, 720 genes associated with breast cancer (BC) were sourced from the DisGeNET database. A Venn diagram analysis was employed to identify the overlapping targets between BC and the genes corresponding to the designed compounds. From this analysis, a total of 25 genes were identified as potential anti-breast cancer (anti-BC) targets and selected for further investigation as key therapeutic targets.
3.6.2. Construction of the compound-target network
The synthesized Thiazolidinone-isoxazoline (TZI) derivative 6 and its regioisomer 6T, along with 25 identified key targets and their corresponding pathways enriched with breast cancer (BC)-related genes, were selected to develop a network diagram depicting the interactions between active compounds, target genes, and associated pathways. The presence of multiple targets for each active compound suggests the potential for a synergistic effect when TZI derivatives are utilized as anti-BC agents. This implies that the interaction of these compounds with various targets may work in concert to enhance their therapeutic efficacy against BC. The overlapping target genes are illustrated in Figure 4S.
3.6.3. Construction of the protein-protein interaction network and identification of pivotal genes
The top 10 overlapping genes were imported into the STRING database to construct the PPI network. In this network, the nodes and their corresponding interactions represent the complex relationships among multiple targets involved in the pathogenesis or progression of the disease. The analysis of the PPI network for the overlapping genes was conducted using a network analyzer. This analysis revealed several key genes, ranked by their degree within the PPI network. The top 10 genes are listed in Table 5 and illustrated in Figure 5S. In the ranking of genes based on their degree scores, EGFR occupies the top position with the highest degree score of 2314, followed closely by PARP1 at 2292. Insulin-like growth factor 1 receptor (IGF1R) and progesterone receptor (PGR) hold the third and fourth ranks with scores of 1730 and 1658, respectively. The list continues with mitogen-activated protein kinase (MAPK14) (1590), prostaglandin-endoperoxide synthase 2 (PTGS2) (1560), and mitogen-activated protein kinase (MAPK8) (1106), demonstrating their significance. Kinase insert domain receptor (KDR), histone deacetylase 6 (HDAC6), and histone deacetylase 2 (HDAC2) complete the top ten, emphasizing their involvement with degree scores of 770, 486, and 480, respectively. This ranking establishes a clear hierarchy of gene significance, with high degree scores indicating strong interconnections among these target genes, particularly within the context of breast cancer pathways. The presence of EGFR and PARP1 at the top underscores their centrality in cellular signaling pathways.
| Rank | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|
| Gene name | EGFR | PARP1 | IGF1R | PGR | MAPK14 | PTGS2 | MAPK8 | KDR | HDAC6 | HDAC2 |
| Degree Score | 2314 | 2292 | 1730 | 1658 | 1590 | 1560 | 1106 | 770 | 486 | 480 |
3.6.4. Network pharmacology analysis of top 10 genes
The network pharmacology analysis identified the top 10 genes ranked by their degree of interaction within the constructed PPI network. The degree score indicates the number of direct interactions each gene has with other genes in the network. Table 5 presents these top-ranked genes along with their respective degree scores.
3.6.5. Role of top 10 genes in cancer
The network pharmacology analysis has highlighted several key genes that play significant roles in cancer. Each of these genes contributes to various aspects of tumor development, progression, and resistance to treatment. The top ten genes play vital roles in the biology and treatment of breast cancer. EGFR is frequently overexpressed in aggressive subtypes, driving proliferation, invasion, and therapy resistance. PARP1 is pivotal in DNA repair pathways, making it a key target for PARP inhibitors in Breast Cancer gene mutation (BRCA)-mutated breast cancers. IGF1R promotes growth and resistance to hormone therapies, contributing to tumor progression. PGR indicates hormone receptor-positive cancers and is targeted in endocrine therapies. MAPK14 supports metastasis through stress responses, while PTGS2 enhances inflammation and tumor microenvironment changes. MAPK8 modulates apoptosis and survival, aiding tumor development. KDR vascular endothelial growth factor receptor 2 (VEGFR2) facilitates angiogenesis, essential for tumor sustenance, while HDAC6 is involved in maintaining estrogen receptor activity, enhancing hormone-driven cancer progression. Lastly, HDAC2 contributes to epigenetic silencing of tumor suppressor genes, further driving malignancy. Together, these genes represent critical pathways and therapeutic targets in breast cancer research and treatment. These genes, which are highly interconnected in cancer pathways, represent prime therapeutic targets for controlling cancer growth and spread.
3.6.6. Kyoto encyclopedia of genes and genomes pathway enrichment analysis
The KEGG pathway enrichment analysis highlights several biological pathways in which the EGFR gene plays an essential role (Figure 6S). In the enrichment plot, EGFR is associated with pathways such as the MAPK signaling pathway, phosphoinositide 3-kinase (PI3K)-Akt signaling pathway, and forkhead box class O (FoxO) signaling pathway, all of which are key regulators of cellular growth and apoptosis. In breast cancer, activation of EGFR leads to persistent stimulation of these downstream pathways, promoting uncontrolled cell division and resistance to programmed cell death. The MAPK signaling pathway in particular facilitates the transmission of proliferative signals from EGFR to the nucleus, driving oncogenic gene expression patterns that contribute to tumor growth. Additionally, pathways such as pathways in cancer, endocrine resistance, and tumor necrosis factor (TNF) signaling pathway also show EGFR involvement. EGFR signaling can interact with hormone receptor pathways, such as those mediated by estrogen receptors, leading to endocrine therapy resistance, a major challenge in treating hormone receptor-positive breast cancers. The TNF signaling pathway links EGFR activity to inflammatory processes within the tumor microenvironment, further enhancing tumor aggressiveness and metastatic potential. Overall, the kyoto encyclopedia of genes and genomes (KEGG) analysis highlights EGFR’s multifaceted role in breast cancer biology. Its enrichment across multiple cancer-related pathways suggests that EGFR acts as a central molecular hub, coordinating diverse oncogenic signals.
3.6.7. Gene ontology (GO) analysis
The GO enrichment analysis provides comprehensive insight into the functional roles of the EGFR gene by categorizing its involvement into three main domains: Biological process (BP), Cellular component (CC), and Molecular function (MF) (Figure 7S). In the biological process category, EGFR is enriched in processes such as signal transduction, positive regulation of cell proliferation, protein phosphorylation, and gene expression regulation. These pathways are central to the initiation and progression of breast cancer. Aberrant activation of EGFR triggers continuous signaling cascades, particularly the MAPK, PI3K-Akt, and janus kinase (JAK) - signal transducer and activator of transcription (STAT) pathways that drive uncontrolled cell growth and inhibit apoptosis. Furthermore, EGFR-mediated phosphorylation events influence transcription factors that promote oncogene expression, leading to enhanced tumor cell survival, angiogenesis, and metastatic potential.
In the cellular component category, EGFR is primarily localized to the plasma membrane, cytoplasm, and cell surface, where it functions as a receptor tyrosine kinase. Upon ligand binding, EGFR undergoes dimerization and autophosphorylation, initiating downstream signaling events. The enrichment in components such as enzyme binding sites, adenosine triphosphate (ATP)-binding regions, and protein complexes highlights EGFR’s role as a signaling hub. In breast cancer cells, this receptor often shows overexpression or mutations, resulting in persistent activation even in the absence of ligands. Such sustained activity at the membrane level reinforces aberrant signaling loops that promote malignancy and resistance to targeted therapies.
Under molecular function, EGFR exhibits strong associations with ATP binding, protein kinase activity, and protein tyrosine kinase activity, all of which are fundamental to its role in phosphorylation-based signaling. These molecular functions enable EGFR to regulate downstream effectors that control proliferation and survival. Additionally, EGFR’s involvement in growth factor binding and enzyme binding reflects its ability to interact with multiple co-receptors and signaling molecules. In breast cancer, particularly in triple-negative and HER2-enriched subtypes, EGFR’s kinase activity contributes to aggressive tumor behavior and therapeutic resistance. The GO enrichment analysis emphasizes that EGFR serves as a multifunctional regulator in breast cancer pathophysiology. These insights further validate EGFR as a potent therapeutic target in the molecular management of breast cancer.
3.6.8. Physicochemical properties
The physico-chemical properties of the designed Thiazolidinone-isoxazole derivatives (6 & 6T) were systematically evaluated to assess their potential as drug candidates against breast cancer. The physicochemical parameters of the designed ligands (compounds 6 and 6T) were evaluated to assess their drug-likeness. Both compounds exhibited a favorable log P value, with 4.70 for compound 6 and 4.39 for 6T, indicating good lipophilicity and potential membrane permeability. The TPSA of 71.07 Å2 for both compounds suggests balanced hydrophilicity, which is suitable for oral bioavailability. Both ligands shared a molecular weight (MW) of 420.54, six hydrogen bond acceptors (HBA), no hydrogen bond donors (HBD), and five rotatable bonds, with no violations of Lipinski’s rule of five. These properties indicate that the compounds are well-suited for further drug development studies (Table 6).
| Product | Log P | TP* | M* | HB* | HBD* | RB* | NV* |
|---|---|---|---|---|---|---|---|
| 6 | 4.70 | 71.07 | 420.54 | 6 | 0 | 5 | 0 |
| 6T | 4.39 | 71.07 | 420.54 | 6 | 0 | 5 | 0 |
TP*: Topological polar surface area, M*: Molecular weight, HB*: Number of hydrogen bond acceptors, HBD*: Number of hydrogen bond donors, RB*: Number of rotatable bonds, NV*: Violations of Lipinski’s rule of five.
3.6.9. Bioactivity
The bioactivity scores of the designed Thiazolidinone-isoxazole derivatives (6 & 6T) against various biological targets are summarized in Table 7. The bioactivity scores for the designed compounds, 6 and 6T, calculated using Molinspiration software, provide a detailed insight into their interaction potential with specific biological targets. These scores are crucial for predicting their functional roles in drug discovery. Compound 6 exhibited moderate activity as an enzyme inhibitor (-0.50) and ion channel modulator (-0.36), suggesting its capability to interact effectively with these targets. While it displayed lower activity as a G protein-coupled receptor (GPCR) ligand (-0.52) and inactive kinase inhibition (-1.12), it still holds promise in pathways involving enzymes and ion channels. Compound 6T showed improved bioactivity across most parameters, particularly with higher scores as a GPCR ligand (-0.47) and enzyme inhibitor (-0.44). Its moderately active profile in these categories indicates a greater likelihood of successful interaction and therapeutic potential. Although kinase inhibition (-1.05) and nuclear receptor ligand activity (-0.55) were less pronounced, the compound’s overall bioactivity profile supports its candidacy for further development. These findings suggest that both compounds have the potential to act as selective modulators, especially in enzymatic and GPCR-mediated pathways. The differences in scores between the compounds highlight structural features that could be optimized to enhance their activity.
| Product | GPCR ligand | Ion channel module | Kinase inhibitor | Nuclear receptor ligand | Protease inhibitor | Enzyme inhibitor |
|---|---|---|---|---|---|---|
| 6 | -0.52 | -0.36 | -1.12 | -0.61 | -0.64 | -0.50 |
| 6T | -0.47 | -0.35 | -1.05 | -0.55 | -0.59 | -0.44 |
3.6.10. Toxicity
The toxicity assessment of the designed Thiazolidinone-isoxazole derivatives (6 and 6T) is summarized in Table 8. The analysis indicates varying toxicity profiles among the compounds. The toxicity assessment of the compounds 6 and 6T was conducted using predictive computational models, revealing distinct safety profiles essential for evaluating their potential as therapeutic agents. Compound 6 was predicted to exhibit hepatotoxicity and immunotoxicity with probabilities of 0.52 and 0.94, respectively, suggesting a moderate risk for liver and immune-related adverse effects. It showed a concerning probability of carcinogenicity (0.67) and mutagenicity (0.50), indicating potential genetic and oncogenic risks. However, the probability of cytotoxicity was relatively low (0.76), suggesting minimal cytotoxic effects. Compound 6T demonstrated a slightly improved toxicity profile with inactive predictions for hepatotoxicity (0.53), immunotoxicity (0.90), and cytotoxicity (0.77). Nevertheless, the active scores for carcinogenicity (0.59) and mutagenicity (0.51) indicate potential risks associated with long-term use. These results highlight areas for structural optimization to mitigate toxicity risks while retaining bioactivity. The predictive toxicity data highlights the importance of balancing therapeutic efficacy with safety considerations in the early stages of drug development.
| Product | Hepatotoxicity | Carcinogenicity | Immunotoxicity | Mutagenicity | Cytotoxicity | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| Toxicity | Prob* | Toxicity | Prob* | Toxicity | Prob* | Toxicity | Prob* | Toxicity | Prob* | |
| 6 | Inactive | 0.52 | Active | 0.67 | Inactive | 0.94 | Active | 0.50 | Inactive | 0.76 |
| 6T | Inactive | 0.53 | Active | 0.59 | Inactive | 0.90 | Active | 0.51 | Inactive | 0.77 |
3.7. Implications and limitations of in silico toxicity predictions
While in silico toxicity models provide a rapid, cost-effective, and ethically favorable means to predict potential adverse effects, they also carry inherent limitations. These models rely on quantitative structure-activity relationships (QSARs), machine learning, and chemical similarity-based algorithms that are trained on existing experimental data. Consequently, their predictive accuracy depends on the quality and diversity of the reference datasets. Moreover, in silico tools cannot fully capture metabolic transformations, bioaccumulation, or species-specific toxicodynamics that influence real-world toxicity. Therefore, these computational findings should be interpreted as preliminary indicators to be validated by experimental toxicological studies.
3.8. Proposed structural modifications to mitigate toxicity
To address the predicted mutagenic and carcinogenic liabilities, several structure-based optimization strategies can be considered. Mutagenicity is often associated with the presence of electrophilic moieties capable of covalently binding to DNA. (1) Modifying or substituting reactive groups such as α, β-unsaturated carbonyls in the Thiazolidinone ring could reduce genotoxic potential. (2) By introducing bulky substituents near reactive sites may limit DNA intercalation or enzymatic activation that leads to mutagenicity. (3) By replacing highly electron-withdrawing substituents (e.g., nitro or halogen groups) with less reactive functional groups can decrease the formation of reactive intermediates responsible for carcinogenicity. And (4) by substituting the isoxazole moiety or modifying the thiazolidinone linker with non-toxic bioisosteres could preserve pharmacological activity while mitigating potential adverse effects. These modifications, coupled with iterative computational re-evaluation, can help refine molecular design to achieve an optimal balance between efficacy and safety.
3.9. Molecular docking
3.9.1. Docking results with EGFR
The molecular docking study was conducted for the two best targets identified from network pharmacology involving EGFR (PDB ID: 4HJO) and PARP1 (PDB ID: 5HA9). The docking results for the designed compounds (6 and 6T) and the standard drug Capivasertib with EGFR (PDB ID: 4HJ0) reveal significant variations in binding affinities and interaction profiles. The docking analysis of the cocrystallized reference ligand with the EGFR protein revealed a binding energy of -7.3 kcal.mol-1, indicating a strong and stable interaction within the active site. The ligand formed hydrogen bonds with key residues Cys773, Met769, and Lys704, contributing significantly to binding stability. Additional interactions included π–σ interactions with Ala719, Leu694, Leu820, and Val702, as well as alkyl interactions with Lys721 and C–H bonding with Gln767 and Pro770. Compound 6 demonstrated the highest docking score of -8.76 kcal.mol-1, indicating a strong binding affinity compared to compound 6T (-8.54 kcal.mol-1) and the standard Capivasertib (-7.07 kcal.mol-1). Compound 6 interacted with 13 amino acid residues within the EGFR active site, forming hydrogen bonds with Lys 851 and Arg 817, van der Waals interactions with Asp 813, and extensive Pi-alkyl interactions with residues such as Lys 721, Leu 834, Leu 838, Phe 699, Ala 840, Ala 847, Tyr 845, Pro 853, and Val 852. These interactions contributed to its strong binding stability. Compound 6T exhibited interactions with 8 residues, including Pi-sulfur interactions with Met 742 and Asp 831, Pi-alkyl interactions with Leu 834, Val 702, Arg 817, Arg 820, and Arg 753, as well as an unfavorable donor-donor interaction with Lys 721, which may have slightly diminished its binding potential. In contrast, the standard drug Capivasertib displayed fewer interactions, forming hydrogen bonds with Asp 813, Pi-alkyl interactions with Val 852 and Ala 847, and van der Waals interactions with Ala 840. These findings indicate that compound 6, with its extensive network of interactions, holds the most promise as an EGFR inhibitor, surpassing both compound 6T and the standard drug. The docking results of the two designed compounds with EGFR target (4HJ0) are shown in Table 9, whilst the 2D and 3D representations of the two compounds alongside the standard drug are displayed in Figure 8.
| Compound | Docking score (Kcal mol−1) | Number of interacting residues | Amino acid and type of interaction |
|---|---|---|---|
| 6 | -8.76 | 13 | Lys 851, Arg 817 (H-bond); Asp 813 (Vanderwaal); Lys 721, Leu 834, 838 Phe 699, Ala 840, 847, 835 Tyr 845, Pro 853, Val 852 (Pi-Alkyl) |
| 6T | -8.54 | 8 | Met 742, Asp 831, (Pi-Sulfur); Lys 721 (Unfavorable donor-donor); Leu 834, Val 702, Arg 817,820, 753 (Pi-Alkyl) |
| STD Capivasertib | -7.07 | 4 | Asp 813 (H-bond); Val 852, Ala 847 (Pi-Alkyl); Ala 840 (Vanderwaals) |
| Cocrystal | -7.3 | 10 | Cys773, Met769, Lys704 (H-bond), Ala719, Leu694, Leu820, Val702 (pi-sigma), Lys721 (alkyl), Gln767, Pro770 (C-H bond) |
- 2D and 3D interactions with EGFR target.
3.9.2. Docking results with PARP1 (PDB ID: 5HA9)
The docking results of the designed compounds (6 and 6T) and the standard drug Capivasertib with the PARP1 target (PDB ID: 5HA9) revealed various types of binding interactions with differing affinities. The docking analysis of the cocrystallized reference ligand with the PARP1 protein exhibited a binding energy of -6.5 kcal.mol-1, suggesting a stable and favorable interaction within the active site. The ligand established a π-π stacking interaction with Tyr246, which plays a crucial role in stabilizing aromatic interactions, along with C-H bonds involving His201 and Arg217. Compound 6 exhibited the strongest binding affinity, with a docking score of -10 kcal.mol-1, interacting with six residues. It formed pi-pi T-shaped interactions with Tyr 235 and Tyr 246, as well as pi-alkyl interactions with Ile 211, His 201, His 248, and Leu 216, indicating robust hydrophobic and aromatic stacking interactions. Compound 6T showed a slightly lower docking score of -8.55 kcal.mol-1 but interacted with eight residues, forming a hydrogen bond with Asp 105, pi-pi T-shaped interactions with Tyr 235, Tyr 246, and His 201, and alkyl interactions with Leu 108, Ala 219, Pro 220, and Arg 217. The additional hydrogen bonding in compound 6T provides a distinct interaction profile compared to compound 6. In contrast, the standard drug Capivasertib demonstrated a docking score of -7.05 kcal.mol-1, interacting with seven residues. It formed hydrogen bonds with Tyr 235, Tyr 246, Glu 327, Glu 102, and Met 229, alongside pi-alkyl interactions involving Leu 324 and Tyr 325. While Capivasertib showcased notable hydrogen bonding, its weaker docking score in comparison to compounds 6 and 6T suggests a relatively lower binding affinity. These findings imply that compound 6, with its superior docking score and interaction profile, holds the most promise as a PARP1 inhibitor. The docking results of the two designed compounds with the EGFR target (4HJ0) are shown in Table 10. Additionally, 2D and 3D representations of the two compounds along with the standard drug are depicted in Figure 9.
| Compound code | Docking score (kcal.mol−1) | Number of interacting residues | Amino acid and type of interaction |
|---|---|---|---|
| 6 | -10 | 6 | Tyr 235, 246(Pi-Pi T shapd); Ile 211, His 201, 248 Leu 216 (Pi-Alkyl) |
| 6T | -8.55 | 8 | Asp 105 (H-bond); Tyr 235, 246, His 201 (Pi-Pi T shaped); Leu 108, Ala 219, Pro 220, Arg 217 (Alkyl) |
| STD capivasertib | -7.05 | 7 | Tyr 235, 246, Glu 327, 102, Met 229 (H-bond); Leu 324, Tyr 325 (Pi-Alkyl) |
| Cocrystal | -6.5 | 3 | Tyr246 (pi-pi stacked), His201, Arg217 (C-H bond) |
- 2D and 3D interactions with PARP1 target.
3.9.3. Theoretical structure-activity relationship (SAR) summary
The comparative in silico SAR analysis between the two regioisomers, compound 6 and compound 6T, reveals that the specific orientation of the isoxazole moiety is a critical determinant of biological activity. Compound 6 demonstrated significantly superior binding affinity (lowest docking scores) for both primary targets (EGFR: -8.76 kcal.mol-1and PARP1: -10.0 kcal.mol-1) compared to compound 6T, a difference directly attributed to its structurally optimal orientation enabling an extensive network of hydrophobic and stabilizing interactions within the active sites, whereas the 6T structure was hindered by unfavorable interactions (e.g., donor-donor interaction with Lys 721 in EGFR). Conversely, the minor structural variation in 6T appeared to slightly enhance its predicted functional scores as a GPCR ligand and enzyme inhibitor, confirming the high sensitivity of the biological profile to regio-chemistry. This theoretical SAR provides a crucial design principle, indicating that the specific regio-orientation found in compound 6 is indispensable for maximizing inhibition of EGFR and PARP1, informing future structural optimization efforts aimed at mitigating predicted toxicity while maintaining high binding affinity.
The theoretical structure-activity relationship (SAR) analysis, based on the comparison of regioisomers 6 and 6T, allowed for the establishment of clear design principles by correlating structure with predicted activities. The specific regio-orientation of the isoxazole ring was identified as the critical determinant: compound 6 shows a significantly higher ΔG binding affinity for the two main targets, EGFR (-8.76 kcal.mol⁻1) and PARP1 (-10.0 kcal.mol⁻1), compared to 6T, its optimal conformation allowing it to generate a vast network of stabilizing interactions (notably Pi-Alkyl interactions in EGFR), while 6T is penalized by unfavorable interactions (e.g., donor-donor with Lys 721 of EGFR). From a mechanistic standpoint, the binding efficiency of compound 6 stems from its positioning in functionally essential regions of enzymes: it targets the ATP-binding site of EGFR and the binding site of PARP1 (notably through pi-pi interactions with Tyr 246), confirming that compound 6 is designed to be a competitive inhibitor that directly blocks the activity of unbound enzymes. Finally, although the 6T regioisomer exhibits lower affinity, its modified structure slightly improves its scores for GPCR ligand and enzyme inhibitor activities, highlighting the high sensitivity of the biological profile to this minimal structural variation.
4. Conclusions
In this study, we investigated the Huisgen cycloaddition to synthesize an isoxazole-thiazolidinone heterocycle from (R)-carvone. The product 6 was purified and characterized by HRMS and NMR. DFT calculations confirmed the high chemoselectivity of the 1,3-dipolar cycloaddition, indicating that the terminal alkyne of product 4 is the most reactive dipolarophile. Theoretical results closely match experimental data. The electronic properties of the synthesized compounds were examined via frontier orbital (HOMO, LUMO) analysis and the Independent gradient model (IGM). Network pharmacology identified EGFR and PARP1 as key targets for breast cancer therapy. In-silico docking studies revealed that compound 6 had the best docking score with low energy conformations of -8.76 kcal.mol⁻1 and -10.00 kcal.mol⁻1 for EGFR and PARP1, respectively. In contrast, compound 6T exhibited weaker binding due to fewer conventional hydrogen bonds.
Acknowledgment
The authors would like to acknowledge Deanship of Graduate Studies and Scientific Research, Taif University for funding this work.
CRediT authorship contribution statement
Saad Alotaibi and Mohammed T Alotaibi: Writing – original draft, visualization, validation, software, methodology, formal analysis, conceptualization.
Declaration of competing interest
There are no conflicts of interest.
Declaration of generative AI and AI-assisted technologies in the writing process
The authors confirm that there was no use of artificial intelligence (AI)-assisted technology for assisting in the writing or editing of the manuscript and no images were manipulated using AI.
Supplementary data
Supplementary material to this article can be found online at https://dx.doi.org/10.25259/AJC_555_2025.
References
- Anticancer and DNA binding studies of potential amino acids based quinazolinone analogs: Synthesis, SAR and molecular docking. Bioorganic Chemistry. 2019;87:252-264. https://doi.org/10.1016/j.bioorg.2019.03.038
- [Google Scholar]
- The significance of N-methylpicolinamides in the development of anticancer therapeutics: Synthesis and structure-activity relationship (SAR) studies. Bioorganic Chemistry. 2019;86:513-537. https://doi.org/10.1016/j.bioorg.2019.02.030
- [Google Scholar]
- In search of definitions: Cancer-associated fibroblasts and their markers. International Journal of Cancer. 2020;146:895-905. https://doi.org/10.1002/ijc.32193
- [Google Scholar]
- Hepatocellular carcinoma in non-alcoholic fatty liver disease: An emerging menace. Journal of Hepatology. 2012;56:1384-1391. https://doi.org/10.1016/j.jhep.2011.10.027
- [Google Scholar]
- Obesity, inflammation, and liver cancer. Journal of Hepatology. 2012;56:704-713. https://doi.org/10.1016/j.jhep.2011.09.020
- [Google Scholar]
- An overview of the pharmacological properties and potential applications of natural monoterpenes. Mini Reviews in Medicinal Chemistry. 2014;14:1156-1168. https://doi.org/10.2174/1389557514666141127145820
- [Google Scholar]
- Biological effects of essential oils--a review. Food and Chemical Toxicology : An International Journal Published for the British Industrial Biological Research Association. 2008;46:446-475. https://doi.org/10.1016/j.fct.2007.09.106
- [Google Scholar]
- Determination of the chemical composition and antioxidant activity of the essential oil of Artemisia dracunculus and of the antifungal and antibacterial activities of Turkish Artemisia absinthium, A. dracunculus, Artemisia santonicum, and Artemisia spicigera essential oils. Journal of Agricultural and Food Chemistry. 2005;53:9452. 9458. https://doi.org/10.1021/jf0516538
- [Google Scholar]
- Beta-thalassemia. Genetics in Medicine. 2010;12:61-76. https://doi.org/10.1097/GIM.0b013e3181cd68ed.
- [Google Scholar]
- Hypotensive and vasorelaxant effects of citronellol, a monoterpene alcohol, in rats. Basic & Clinical Pharmacology & Toxicology. 2010;106:331-337. https://doi.org/10.1111/j.1742-7843.2009.00492.x
- [Google Scholar]
- New thiazole-2(3H)-thiones containing 4-(3,4,5-trimethoxyphenyl) moiety as anticancer agents. European Journal of Medicinal Chemistry. 2020;185:111784. https://doi.org/10.1016/j.ejmech.2019.111784
- [Google Scholar]
- The recent progress of isoxazole in medicinal chemistry. Bioorganic & Medicinal Chemistry. 2018;26:3065-3075. https://doi.org/10.1016/j.bmc.2018.05.013
- [Google Scholar]
- Synthesis and biological evaluation of chalcone incorporated thaizole-isoxazole derivatives as anticancer agents. Chemical Data Collections. 2025;55:101177. https://doi.org/10.1016/j.cdc.2024.101177
- [Google Scholar]
- Synthesis and X-ray structure of potent anticancer 4-(4-methoxyphenyl)-3-(3,4,5-trimethoxyphenyl)isoxazole. Mendeleev Communications. 2024;34:511-513. https://doi.org/10.1016/j.mencom.2024.06.013
- [Google Scholar]
- Novel nitric oxide-donating compound (S,R)-3-phenyl-4,5-dihydro-5-isoxazole acetic acid–nitric oxide (GIT-27NO) induces p53 mediated apoptosis in human A375 melanoma cells. Nitric Oxide. 2008;19:177-183. https://doi.org/10.1016/j.niox.2008.04.004
- [Google Scholar]
- The synthetic and therapeutic expedition of isoxazole and its analogs. Medicinal Chemistry Research : An International Journal for Rapid Communications on Design and Mechanisms of Action Of Biologically Active Agents. 2018;27:1309-1344. https://doi.org/10.1007/s00044-018-2152-6
- [Google Scholar]
- Prediction of new Hsp90 inhibitors based on 3,4-isoxazolediamide scaffold using QSAR study, molecular docking and molecular dynamic simulation. Daru : Journal of Faculty of Pharmacy, Tehran University of Medical Sciences. 2017;25:17. https://doi.org/10.1186/s40199-017-0182-0
- [Google Scholar]
- Convenient synthesis of highly functionalized isoxazoles including an amidine skeleton based on trichloroacetamidine, alkynes, and hydroxyimidoyl chloride. Tetrahedron Letters. 2025;155:155443. https://doi.org/10.1016/j.tetlet.2024.155443
- [Google Scholar]
- Anticancer effects of abietane diterpene 7α-Acetoxy-6β-hydroxyroyleanone from plectranthus grandidentatus and Its semi-synthetic analogs: An in silico computational approach. Molecules (Basel, Switzerland). 2024;29:1807. https://doi.org/10.3390/molecules29081807
- [Google Scholar]
- Erlotinib in previously treated non–small-cell lung cancer. The New England Journal of Medicine. 2005;353:123-132. https://doi.org/10.1056/NEJMoa050753
- [Google Scholar]
- Gefitinib or chemotherapy for non–small-cell lung cancer with mutated EGFR. New England Journal of Medicine. 2010;362:2380-2388. https://doi.org/10.1056/nejmoa0909530
- [Google Scholar]
- Effect of external stresses on protein conformation: A computer modelling study. European Biophysics Journal : EBJ. 2004;33:121-129. https://doi.org/10.1007/s00249-003-0359-y
- [Google Scholar]
- Rational design, synthesis and anticancer screening of 1,2,4-oxadiazole incorporated thieno[2,3-d]thiazole-isoxazole-pyridine derivatives. Results in Chemistry. 2024;11:101823. https://doi.org/10.1016/j.rechem.2024.101823
- [Google Scholar]
- Design, synthesis and biological evaluation of aryl isoxazole incorporated pyrimidin-2-yl)oxazolo[4,5-b]pyridine derivatives as anticancer agents. Results in Chemistry. 2024;9:101628. https://doi.org/10.1016/j.rechem.2024.101628
- [Google Scholar]
- Design and synthesis of 3,5-disubstituted isoxazoles by Cu-mediated 1,3-dipolar cycloaddition and their in silico evaluation as potential GABAB receptor modulators. Tetrahedron. 2024;168:134336. https://doi.org/10.1016/j.tet.2024.134336
- [Google Scholar]
- SwissTargetPrediction: Updated data and new features for efficient prediction of protein targets of small molecules. Nucleic Acids Research. 2019;47:W357-W364. https://doi.org/10.1093/nar/gkz382
- [Google Scholar]
- SwissTargetPrediction: A web server for target prediction of bioactive small molecules. Nucleic Acids Research. 2014;42:W32-W38. https://doi.org/10.1093/nar/gku293
- [Google Scholar]
- Towards a universal SMILES representation – A standard method to generate canonical SMILES based on the InChI. J Cheminformatics 4. 2012;22 https://doi.org/10.1186/1758-2946-4-22
- [Google Scholar]
- MalaCards: An amalgamated human disease compendium with diverse clinical and genetic annotation and structured search. Nucleic Acids Research. 2017;45:D877-D887. https://doi.org/10.1093/nar/gkw1012
- [Google Scholar]
- InteractiVenn: A web-based tool for the analysis of sets through Venn diagrams. BMC Bioinformatics. 2015;16:169. https://doi.org/10.1186/s12859-015-0611-3
- [Google Scholar]
- DAVID Knowledgebase: a gene-centered database integrating heterogeneous gene annotation resources to facilitate high-throughput gene functional analysis. BMC Bioinformatics. 2007;2:426. https://doi.org/ 10.1186/1471-2105-8-426.
- [Google Scholar]
- Straightforward synthesis and in silico evaluation of pyrazolylthiazolidinone derivatives as prospective antiproliferative agents. Bioorganic Chemistry. 2025;165:109036. https://doi.org/10.1016/j.bioorg.2025.109036
- [Google Scholar]
- Construction of miRNA-mRNA network for the identification of key biological markers and their associated pathways in IgA nephropathy by employing the integrated bioinformatics analysis. Saudi Journal of Biological Sciences. 2021;28:4938-4945. https://doi.org/10.1016/j.sjbs.2021.06.079
- [Google Scholar]
- The STRING database in 2021: Customizable protein–protein networks, and functional characterization of user-uploaded gene/measurement sets. Nucleic Acids Research. 2021;49:D605-D612. https://doi.org/10.1093/nar/gkaa1074
- [Google Scholar]
- Thymol-1,2,3-triazole derivatives: Network pharmacology, molecular simulations and synthesis targeting breast cancer. Journal of Molecular Structure. 2025;1321:140060. https://doi.org/10.1016/j.molstruc.2024.140060
- [Google Scholar]
- Exploring the cyclization of thiosemicarbazone to 1,3,4-thiadiazole: Synthesis, characterization and in-silico study. Journal of Molecular Structure. 2025;1322:140385. https://doi.org/10.1016/j.molstruc.2024.140385
- [Google Scholar]
- Molinspiration cheminformatics free web services, Slovensky Grob, Slovakia. https://www.molinspiration.com https://www.molinspiration.com.
- Experimental and computational approaches to estimate solubility and permeability in drug discovery and development settings. Advanced Drug Delivery Reviews. 2001;46:3-26. https://doi.org/10.1016/s0169-409x(00)00129-0
- [Google Scholar]
- Exploring cyclin-dependent kinase inhibitors: A comprehensive study in search of CDK-6 inhibitors using a pharmacophore modelling and dynamics approach. RSC Advances. 2023;13:33770-33785. https://doi.org/10.1039/d3ra05672d
- [Google Scholar]
- ProTox-II: A webserver for the prediction of toxicity of chemicals. Nucleic Acids Research. 2018;46:W257-W263. https://doi.org/10.1093/nar/gky318
- [Google Scholar]
- Synthesis of novel N4-substituted and C2-disubstituted 1,4-benzothiazine-1,1-dioxide derivatives: Integrative computational strategies for breast cancer therapy. Journal of Molecular Structure. 2025;1338:142310. https://doi.org/10.1016/j.molstruc.2025.142310
- [Google Scholar]
- Anti-bacterial activity of furan chalcone derivatives against mycobacterium tuberculosis: Design, synthesis, anti-bacterial screening, pharmacokinetic properties, and toxicity parameters. Current microbiology. 2025;82:373. https://doi.org/10.1007/s00284-025-04319-6
- [Google Scholar]
- Targeting aurora A kinase: Computational discovery of potent inhibitors through integrated pharmacophore and simulation approaches. Computational Biology and Chemistry. 2026;120:108690. https://doi.org/10.1016/j.compbiolchem.2025.108690
- [Google Scholar]
- Autodock4 and AutoDockTools4: Automated docking with selective receptor flexiblity. Journal of Computational Chemistry. 2009;16:2785-2791. https://doi.org/10.1002/jcc.21256
- [Google Scholar]
- UCSF chimera--a visualization system for exploratory research and analysis. Journal of Computational Chemistry. 2004;25:1605-1612. https://doi.org/10.1002/jcc.20084
- [Google Scholar]
- A novel brominated chalcone derivative as a promising multi-target inhibitor against multidrug-resistant listeria monocytogenes. Microbial Pathogenesis. 2024;196:106968. https://doi.org/10.1016/j.micpath.2024.106968
- [Google Scholar]
- UCSF Chimera a visualization system for exploratory research and analysis. Journal of Computational Chemistry. 2004;25:1605-1612. https://doi.org/10.1002/jcc.20084
- [Google Scholar]
- Synthesis, antitubercular evaluation, molecular docking study, and teratogenicity studies of novel triazolo quinazoline hybrids. Russian Journal of General Chemistry. 2023;93:2684-2693. https://doi.org/10.1134/s1070363223100237
- [Google Scholar]
- 2,4-substituted oxazolones: Antioxidant potential exploration. Journal of Young Pharmacists. 2024;16:244-251. https://doi.org/10.5530/jyp.2024.16.32
- [Google Scholar]
- Multifaceted analysis of novel trisubstituted imidazole chromen derivatives : Design, synthesis, characterization, molecular docking, and antimicrobial evaluation. Journal of Molecular Structure. 2025;1346:143185. https://doi.org/10.1016/j.molstruc.2025.143185
- [Google Scholar]
- In silico design and synthesis of some new imidazole derivatives for tuberculosis. Indian Journal of Heterocyclic Chemistry. 2023;33:43. https://doi.org/10.59467/ijhc.2023.33.43
- [Google Scholar]
- Unravelling the strain-promoted [3+2] cycloaddition reactions of phenyl azide with cycloalkynes from the molecular electron density theory perspective. New Journal of Chemistry. 2020;44:13633-13643. https://doi.org/10.1039/d0nj02711a
- [Google Scholar]
- The nucleophilicity N index in organic chemistry. Organic & Biomolecular Chemistry. 2011;9:7168-7175. https://doi.org/10.1039/c1ob05856h
- [Google Scholar]
- A molecular electron density theory study of the grignard reagent‐mediated regioselective direct synthesis of 1,5‐disubstituted‐1,2,3‐triazoles. Journal of Physical Organic Chemistry. 2020;33 https://doi.org/10.1002/poc.4062
- [Google Scholar]