Identification of chrysin derivatives anticancer potential in MCF7 cells: biological insights and in silico evaluation

ABSTRACT

Aim

To synthesize and evaluate the anticancer potential and mechanism of a naturally occurring compound, chrysin derivatives.

Materials & Methods

A series of 7-substituted phenyl pyrano derivatives of chrysin (3a–k) were synthesized by Michael-type addition reaction and their structures were elucidated using spectroscopic techniques, such as FT-IR, H¹ NMR, C¹³ NMR, and MS. In vitro anticancer and cytotoxicity effects were evaluated using MCF7 and mesenchymal stem cells (MSCs). Apoptosis mechanism was evaluated through the expression of pro- and anti-apoptotic proteins, for instance, Bax, Bcl-2, p53, and p21, and the binding score and stability was computed using AutoDock Vina and GROMACS. In silico ADMET analysis was performed via web-based tools like Swiss ADME, pkCSM, ADMETlab 2.0, PreADMET, ProTox II, and Molinspiration.

Results

Structure activity relationship (SAR) analysis revealed that the 4-hydroxy substituted phenyl derivative (3h) is important for anticancer activity. 3 h enhanced the expression of Bax, Bcl-2, and p53 while decreases in the expression of oncogene p21 at 16.5 µM concentration showed superior activity to standard carboplatin and was found safe up to 77.95 μM. All the derivatives displayed favorable pharmacokinetic and drug-like properties.

Conclusion

The 4-hydroxy substituted phenyl derivative (3h) spectacled enhanced anticancer and safety profile along with considerable pharmacokinetic parameters.

GRAPHICAL ABSTRACT

1. Introduction

Cancer cases has globally risen to 8.1 million annually [1], with breast cancer accounting for 25% of all malignancies and ranks as the fifth leading cause of mortality. Surgery is considered the primary treatment for breast cancer while adjuvant treatment includes radiation and systemic therapy for instance chemotherapeutic, hormonal, and biological agents for metastasized breast cancer [2]. In cancer, the balance between cell proliferation and apoptosis, programmed cell death, is disrupted. Cancer cells often exhibit increased proliferation and reduced apoptosis, leading to uncontrolled tumor growth [3,4]. The intrinsic and extrinsic pathways are primarily initiate apoptosis: intrinsic pathway is activated by cellular stress, DNA damage, or oncogene activation which involves the release of cytochrome c proteins from the mitochondria [5]. The Bcl-2 family protein plays a crucial role in regulating this pathway, with some members promoting pro-apoptotic protein and inhibiting anti-apoptotic proteins [6]. Bcl-2 is an anti-apoptotic protein that is upregulated frequently across multiple cancer types. Contrarily, the protein Bax is involved in pro-apoptotic pathways and induce apoptosis [7]. Another important apoptotic protein p53 controls the production of downstream target genes [8–10] such as cyclin G [11] and p21 [12]. Lowered expression of p53 is associated with decreased apoptosis, thereby facilitating tumor growth and progression [13]. The p21 significantly inhibits cyclin-dependent kinases (CDKs), and involves in DNA repair, cellular differentiation, and apoptosis [14,15].

Many natural compounds exhibit anticancer activity, including the well-known drugs paclitaxel, vincristine, vinblastine, etoposide, and topotecan [16]. Chrysin is 5,7-dihydroxyflavone exists in glycosylated forms in a variety of plant species, including Desmos cochinchinensis [17], Mangifera indica [18], Oroxylum indicum [19], Alpina oxyphylla [20], Eriodictyon califomicum [21], and Scutellaria sp [22]. It is also found in marine endophytic strains [23], honey [24,25], propolis [26], mushrooms [27] and stingless bee products [27]. Chrysin has a variety of biological potentials for instance anti-allergic [28], antibacterial, antifungal [29], antiviral [30], anti-inflammatory [31,32], antidepressant [33], antidiabetic [34], antihypertensive [35], antioxidant [36], and anticancer properties [37].

Chrysin induces apoptosis and inhibits MCF7 cells it can be a promising candidate or lead compound for anticancer therapy [38,39]. It increases the expression of Bax, Bad, and Bak while simultaneously decreasing that of Bcl-2, resulting in mitochondrial-driven apoptotic signaling and caspase activation across multiple cancer models [40,41]. Chrysin treatment upregulates p53 and p21 expression via the ATM-Chk2 pathway in MCF7 cells, arresting the cell cycle without causing DNA damage [42] the upregulation of p21 May occur through p53-independent mechanisms in other cancer cell lines [41,43]. Chrysin exhibits considerable in vitro action only however it’s in vivo activity retards owing to its low oral bioavailability [44–46] for the account of quick metabolism, low permeability, and poor solubility. Derivatization is indeed a viable approach to enhance the physicochemical properties of chrysin while preserving its biological activity. By modifying its chemical structure, researchers can improve chrysin’s solubility, bioavailability, and other properties, potentially leading to better therapeutic outcomes [46–48]. Various chemical derivatives were synthesized to improve its physical and anticancer properties including pyrimidine [49], 1, 2, 3-triazole [50] and de-allyl PAC-1 hybrid analogues [51]. The amino acid- [52] and urea [53]-based chrysin derivatives increased Bax activation and suppressed Bcl-2 protein expression in a xenograft model of breast cancer with a satisfactory safety profile. A synthetic complex of chrysin with ruthenium displayed potent anti-breast cancer efficacy against MCF7 cells by elevating both Bax and p53, downregulating Bcl-2 in the xenograft model [54], demonstrating the translational potential of compounds derived from chrysin in breast cancer treatment.

This study aims to synthesize 7-substituted phenyl pyrano derivatives of chrysin to target the key regulators apoptotic proteins including Bax, Bcl-2, p53, and p21 to treat breast cancer among them, compound 3 h demonstrated higher potency than chrysin and comparable potency to carboplatin. Furthermore, 3 h was evaluated for its protein-binding interactions using molecular docking in AutoDock Vina, and the stability of the subsequent binding complex was examined through molecular dynamics (MD) simulation in GROMACS.

2. Materials and method

The material used for the synthesis of chrysin derivatives; chrysin powder (≥99%), substituted benzaldehyde (≥99%), malononitrile (≥99%), and piperidine (99%). All reactions were carried out in a round bottom that had been oven dried to 110°C. The solvents and reagents used were purchased commercially and utilized without additional purification. Spot visualization was carried out under UV illumination (254 nm) to evaluate the reaction development using thin layer chromatography (TLC) on MERK silica gel plates using ethyl acetate: n-hexane (7:3) as a mobile phase. The melting points were determined using the Stuart melting point apparatus (SMP 30) and were not adjusted. The H¹ (400 MHz) and C¹³ (100 MHz) NMR spectra were recorded on Bucker instruments in CDCl₃, employing TMS as a reference for reporting chemical changes in δ ppm. Infrared spectra were collected between 4000 and 500 cm⁻¹ using a Shimadzu FT-IR spectrophotometer. Mass spectra (EI+) were acquired from Finnigan MAT-321A. Synthesis of derivatives (3a-k) was done by following the reported procedure.

2.1. Chemistry

2.1.1. General procedure for the synthesis of chrysin derivatives (3a-k)

1.2 mmol of a particular aromatic aldehyde and 1.2 mmol (79.272 mg) of malononitrile solutions were prepared in 5 ml of ethanol. At room temperature, both solutions were then added to a 10% solution of piperidine. The mixture was then heated to 50°C and stirred for five minutes, after which 1.0 mmol (254 mg) of chrysin was added. The reaction mixture was agitated under reflux, and TLC was used to track the progress of the reaction using 7:3 ethyl acetate: n-hexane as a solvent system. When TLC indicates the reaction has been completed, the mixture is allowed to cool at room temperature. Filtration was employed to separate the developed precipitate, and reduced pressure was used to concentrate the filtrate. The resultant precipitates were washed with ethanol before being dried in a desiccator.

2.1.2. 8-amino-5-hydroxy-4-oxo-2,10-diphenyl-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3a)

Yield: 80%; M.P. 288–290°C; FT-IR (cm⁻¹): 3710 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1700 cm⁻¹ (C = O), 1850–1720 cm⁻¹ (C = C, aromatic), 1520–1500 cm⁻¹ (C = C, cyclic), 1460 cm⁻¹ (C-H), 1085–1010 cm⁻¹ (C-O), 735–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 4.67 (s, 2 H, NH₂), 4.88 (s, 1 H, CH), 6.69 (s, 1 H, H_Ar), 6.74 (s, 1 H, H_Ar), 7.38–7.39 (m, 3 H, H_Ar), 7.54–7.61 (m, 4 H, H_Ar), 7.89–7.93 (m, 3 H, H_Ar), 13.10 (s, 1 H, OH); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 35.91 (CH), 94.64, 105.88, 107.56 (C = C), 126.40, 127.21, 127.50, 128.62, 129.20, 129.28, 132.21 (C = C_Ar), 151.86 ( = C-O), 182.52 (C = O); EI-MS: m/z [M] calc. For C₂₅H₁₆N₂O₄: 408.11; found [M]⁺ 408.2, [M + H]⁺ 409.2

2.1.3. 8-amino-5-hydroxy-10-(2-hydroxyphenyl)-4-oxo-2-phenyl-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3b)

Yield: 74%; M.P. 293–294°C; FT-IR (cm⁻¹): 3700 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1770 cm⁻¹ (C = O), 1710–1640 cm⁻¹ (C = C, aromatic), 1530–1500 cm⁻¹ (N-O), 1460 cm⁻¹ (C-H), 1140 cm⁻¹ (C-O), 740–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 4.31 (s, 2H, NH₂), 4.98 (s, 1H, CH), 6.56 (s, 1H, HAr), 6.69 (s, 1H, H_Ar), 6.96 (s, 1H, H_Ar), 7.07–7.09 (d, 2H, J = 8 Hz, H_Ar), 7.46–7.51 (m, 4H, H_Ar), 7.61 (d, 2H, J = 4.5 Hz, H_Ar), 7.81 (d, 2H, J = 4.7 Hz, H_Ar), 12.74 (s, 1H, OH), 13.36 (s, 1H, OH); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 26.13 (CH), 49.56 (C-CN), 97.53, (C = C), 117.12, 117.56, 118.84 (C = C_Ar), 119.85 (CΞN), 124.19, 124.41, 126.57, 128.22, 129.22, 130.57, 131.76, 132.84, 137.25 (C = C_Ar), 150.67, 160.44, 161.84, 164.93, 165.24, 166.21 ( = C-O); EI-MS: m/z [M] calc. For C₂₅H₁₆N₂O₅: 424.11; found [M]⁺ 424.3, [M + H]⁺ 425.0

2.1.4. 8-amino-10-(2-chlorophenyl)-5-hydroxy-4-oxo-2-phenyl-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3c)

Yield: 82%; M.P. 297–299°C; FT-IR (cm⁻¹): 3700 cm⁻¹ (O-H), 3620–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1770 cm⁻¹ (C = O), 1710–1640 cm⁻¹ (C = C, aromatic), 1530–1500 cm⁻¹ (N-O), 1460 cm⁻¹ (C-H), 1085–1010 cm⁻¹ (C-O), 800 cm⁻¹ (C-Cl); H¹ NMR (400 MHz, CDCl₃, δ ppm): 6.93 (s, 1H, H_Ar), 6.99 (s, 1H, H_Ar), 7.10–7.12 (m, 2H, H_Ar), 7.18 (s, 1H, H_Ar), 7.33–7.37 (m, 4H, H_Ar), 7.46 (m, 1H, H_Ar), 8.43–8.45 (d, 2H, J = 8 Hz, H_Ar); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 26.13 (CH), 49.56 (C-CN), 97.80 (C = C), 117.12, 117.56, 118.84 (C = C_Ar), 119.85 (CΞN), 124.19, 124.41, 126.57, 128.22, 129.22, 130.57, 131.76, 132.84, 137.25 (C = C_Ar), 150.67, 160.44, 161.84, 164.93, 165.24, 166.21 ( = C-O); EI-MS: m/z [M] calc. For C₂₅H₁₅ClN₂O₄: 442.07; found [M]⁺ 442.1, [M + H]⁺ 443.2

2.1.5. 8-amino-5-hydroxy-4-oxo-2-phenyl-10-(pyridin-2-yl)-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3d)

Yield: 73%; M.P. 273–275°C; FT-IR (cm⁻¹): 3700 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1770 cm⁻¹ (C = O), 1710–1640 cm⁻¹ (C = C, aromatic), 1530–1500 cm⁻¹ (N-O), 1460 cm⁻¹ (C-H), 1140 cm⁻¹ (C-O), 740–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 5.17 (s, 2H, NH₂), 6.51 (s, 1H, H_Ar), 6.59 (s, 1H, H_Ar), 7.20 (s, 1H, H_Ar), 7.50–7.55 (m, 5H, H_Ar), 7.65 (s, 1H, H_Ar), 7.86–7.88 (d, 2H, J =8 Hz, H_Ar), 8.62 (s, 1H, H_Ar), 13.15 (s, 1H, OH); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 23.92, 25.67 (CH), 52.75 (C=C), 69.28 (C-CN), 95.11, 104.04, 105.46, 107.78 (C=C), 122.75 (CΞN), 123.66, 126.23, 129.03, 131.50, 131.63, 136.88 (C=C_Ar), 149.64 158.65 (C_Ar = N), 157.32, 159.33, 163.57, 165.96 ( = C-O), 182.24 (C = O); EI-MS: m/z [M] calc. For C₂₄H₁₅N₃O₄: 409.11; found [M]⁺ 408.2, [M + H]⁺ 409.2

2.1.6. 8-amino-10-(2-chloro-3-methoxyphenyl)-5-hydroxy-4-oxo-2-phenyl-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3e)

Yield: 87%; M.P. 291–293°C; FT-IR (cm⁻¹): 3700 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1770 cm⁻¹ (C = O), 1710–1640 cm⁻¹ (C = C, aromatic), 1530–1500 cm⁻¹ (N-O), 1460 cm⁻¹ (C-H), 1085–1010 cm⁻¹ (C-O), 800 cm⁻¹ (C-Cl) 740–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 3.91 (s, 3H, CH₃), 4.67 (s, 2H, NH₂), 5.40 (s, 1H, CH), 6.68 (s, 1H, H_Ar), 6.72 (s, 1H, H_Ar), 6.79–6.81 (d, 2H, J = 8 Hz, H_Ar), 6.83–6.85 (d, 2H, J = 8 Hz, H_Ar), 7.15–7.18 (m, 1H, H_Ar), 7.55–7.63 (m, 3H, H_Ar), 7.89–7.91 (d, 2H, J = 8 Hz, H_Ar), 12.98 (s, 1H, OH); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 29.70 (CH), 56.23 (O-CH₃), 58.50 (C-CN), 94.36, 105.91 (C = C), 110.61 ( = C-Cl), 119.85 (CΞN), 126.39, 127.13, 129.20, 132.20 (C = C_Ar), 154.02, 159.27 ( = C-O); EI-MS: m/z [M] calc. For C₂₆H₁₇ClN₂O₅: 472.08; found [M]⁺ 472.2, [M + H]⁺ 473.0

2.1.7. 8-amino-5-hydroxy-10-(3-nitrophenyl)-4-oxo-2-phenyl-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3f)

Yield: 96%; M.P. 295–297°C; FT-IR (cm⁻¹): 3700 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1770 cm⁻¹ (C = O), 1710–1640 cm⁻¹ (C = C, aromatic), 1530–1500 cm⁻¹ (N-O), 1460 cm⁻¹ (C-H), 1085–1010 cm⁻¹ (C-O), 735–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 4.80 (s, 2H, NH₂), 4.90 (s, 1H, CH), 6.56 (s, 1H, H_Ar), 6.67 (s, 1H, H_Ar), 7.40–7.52 (m, 3H, H_Ar),), 7.65–7.66 (d, 2H, J = 5 Hz, H_Ar), 7.80–7.84 (m, 3H, H_Ar), 7.97 (s, 1H, H_Ar), 8.02 (d, 2H, J = 6.7 Hz, H_Ar), 13.08 (s, 1H, OH); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 21.10, 35.51 (CH), 62.79 (C-CN), 94.63, 105.85, 107.71, 107.87 (C = C), 119.09 (CΞN), 126.40, 127.34, 129.19, 129.27, 129.34, 130.96, 132.20, 136.83, 140.48 (C = C_Ar), 158.33, 159.06, 164.64 ( = C-O), 182.50 (C = O); EI-MS: m/z [M] calc. For C₂₅H₁₅N₃O_6: 453.10; found [M]⁺ 453.3, [M + H]⁺454.3

2.1.8. 8-amino-10-(3,4-dimethoxyphenyl)-5-hydroxy-4-oxo-2-phenyl-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3 g)

Yield: 75%; M.P. 281–283°C; FT-IR (cm⁻¹): 3700 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1770 cm⁻¹ (C = O), 1710–1640 cm⁻¹ (C = C, aromatic), 1530–1500 cm⁻¹ (N-O), 1460 cm⁻¹ (C-H), 1450–1380 cm⁻¹ (C-H), 1085 cm⁻¹ (C-O), 740–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 3.85 (s, 6H, CH₃), 4.68 (s, 2H, NH₂), 4.83 (s, 1H, CH), 6.69 (s, 1H, H_Ar), 6.72 (s, 1H, H_Ar), 6.75–6.83 (m, 3H, H_Ar), 6.87 (s, 1H, H_Ar), 7.55–7.60 (m, 3H, H_Ar), 7.91–7.93 (d, 2H, J = 7.6 Hz, H_Ar), 13.16 (s, 1H, OH); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 35.47 (CH), 55.87, 55.94 (O-CH₃), 62.71 (C-CN), 94.62, 105.86, 107.73, 107.89 (C = C), 111.12, 111.22, 119.13 (C = C), 119.31 (CΞN), 126.40, 129.21, 130.94, 132.23 (C = CAr), 148.17, 148.95, 153.70, 155.69, 158.42, 159.10, 164.69 ( = C-O), 182.55 (C = O); EI-MS: m/z [M] calc. For C₂₇H₂₀N₂O_6: 468.13; found [M]⁺ 468.1, [M + H]⁺ 469.2

2.1.9. 8-amino-5-hydroxy-10-(4-hydroxyphenyl)-4-oxo-2-phenyl-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3 h)

Yield: 70%; M.P. 296–298°C; FT-IR (cm⁻¹): 3700 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1770 cm⁻¹ (C = O), 1710–1640 cm⁻¹ (C = C, aromatic), 1530–1500 cm⁻¹ (N-O), 1460 cm⁻¹ (C-H), 1140 cm⁻¹ (C-O), 740–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 4.31 (s, 2H, NH₂), 4.98 (s, 1H, CH), 6.56 (s, 1H, HAr), 6.69 (s, 1H, H_Ar), 6.96 (s, 1H, H_Ar), 7.07–7.09 (d, 2H, J = 8 Hz, H_Ar), 7.46–7.51 (m, 4H, H_Ar), 7.61 (d, 2H, J = 4.5 Hz, H_Ar), 7.81 (d, 2H, J = 4.7 Hz, H_Ar), 12.74 (s, 1H, OH), 13.36 (s, 1H, OH); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 26.13 (CH), 49.56 (C-CN), 97.53, (C = C), 117.12, 117.56, 118.84 (C = C_Ar), 119.85 (CΞN), 124.19, 124.41, 126.57, 128.22, 129.22, 130.57, 131.76, 132.84, 137.25 (C = C_Ar), 150.67, 160.44, 161.84, 164.93, 165.24, 166.21 ( = C-O); EI-MS: m/z [M] calc. For C₂₅H₁₆N₂O₅: 424.11; found [M]⁺ 424.3, [M + H]⁺ 425.0

2.1.10. 8-amino-5-hydroxy-10-(4-isopropylphenyl)-4-oxo-2-phenyl-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3i)

Yield: 72%; M.P. 300–302°C; FT-IR (cm⁻¹): 3700 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1770 cm⁻¹ (C = O), 1710–1640 cm⁻¹ (C = C, aromatic), 1530–1500 cm⁻¹ (N-O), 1460 cm⁻¹ (C-H), 1450–1380 cm⁻¹ (C-H), 1150–1020 cm⁻¹ (C-O), 740–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 1.22–1.24 (d, 8H, J = 8 Hz, CH₃), 2.64 (s, 1H, CH), 4.69 (s, 2H, NH₂), 4.85 (s, 1H, CH), 6.69 (s, 1H, H_Ar), 6.73 (s, 1H, H_Ar), 7.15–7.22 (m, 5H, H_Ar), 7.54–7.61 (m, 3H, H_Ar), 7.89–7.90 (d, 2H, J = 7.5 Hz, H_Ar), 13.09 (s, 1H, OH); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 23.90, 33.69, 35.43 (CH), 40.96 (C-CN), 94.63 (C = C), 126.40, 126.67, 127.30, 129.20, 132.20 (C = C_Ar); EI-MS: m/z [M] calc. For C₂₈H₂₂N₂O_4: 450.16; found [M]⁺ 450.0, [M + H]⁺ 451.3

2.1.11. 8-amino-5-hydroxy-4-oxo-2-phenyl-10-(p-tolyl)-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3j)

Yield: 71%; M.P. 286–288°C; FT-IR (cm⁻¹): 3710 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1700 cm⁻¹ (C = O), 1850–1720 cm⁻¹ (C = C, aromatic), 1520–1500 cm⁻¹ (C = C, cyclic), 1460 cm⁻¹ (C-H), 1430 cm⁻¹ (C-H), 1085–1010 cm⁻¹ (C-O), 735–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 2.32–2.34 (d, 3H, J = 6 Hz, CH₃), 4.65 (s, 2H, NH₂), 4.85 (s, 1H, CH), 6.69 (s, 1H, H_Ar), 6.73 (s, 1H, H_Ar), 7.12–7.13 (d, 2H, J = 6.7 Hz, H_Ar), 7.17–7.19 (d, 2H, J = 6.9 Hz, H_Ar), 7.54–7.62 (m, 3H, H_Ar), 7.89–7.91 (d, 2H, J = 8 Hz, H_Ar), 13.09 (s, 1H, OH); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 21.10, 35.51 (CH), 62.83 (C-CN), 94.63 (C = C), 105.86, 107.70 (C = C_Ar), 119.08 (CΞN), 126.40, 127.34, 128.53, 129.20, 130.97, 132.20, 136.83, 140.47 (C = C_Ar), 153.74, 155.70, 158.32, 159.07, 164.64 ( = C-O), 182.50 (C = O); EI-MS: m/z [M] calc. For C₂₆H₁₈N₂O₄: 422.13; found [M]⁺ 422.3, [M + H]⁺ 423.5

2.1.12. 8-amino-10-(4-(dimethylamino)phenyl)-5-hydroxy-4-oxo-2-phenyl-4 H,10 H-pyrano[2,3-f]chromene-9-carbonitrile (3k)

Yield: 67%; M.P. 259–261°C; FT-IR (cm⁻¹): 3700 cm⁻¹ (O-H), 3600–3500 cm⁻¹ (N-H), 2380 cm⁻¹ (CΞN), 1770 cm⁻¹ (C = O), 1710–1640 cm⁻¹ (C = C, aromatic), 1530–1500 cm⁻¹ (N-O), 1465 cm⁻¹ (C-H), 1450–1380 cm⁻¹ (C-H), 1140 cm⁻¹ (C-O), 740–675 cm⁻¹ (C = C); H¹ NMR (400 MHz, CDCl₃, δ ppm): 3.09 (s, 6H, CH₃), 6.71 (s, 2H, NH₂), 6.73 (s, 1H, H_Ar), 6.97 (s, 1H, H_Ar), 7.48 (s, 1H, H_Ar), 7.82–7.84 (d, 2H, J = 8 Hz, H_Ar); C¹³ NMR (100 MHz, CDCl₃, δ ppm): 40.13 (CH), 72.03 (C-CN), 111.66, 114.90, 115.97 (C = C_Ar), 119.38 (CΞN), 133.80 (C = C_Ar), 154.21, 158.10 ( = C-O); EI-MS: m/z [M] calc. For C₂₇H₂₁N₃O_4: 451.15; found [M]⁺ 450.0, [M + H]⁺ 451.3

2.2. In vitro study

2.2.1. Cell line and cell culture conditions

Human breast cancer cell line (MCF7) and umbilical cord mesenchymal stem cells (MSCs) were used for cytotoxic investigation. The MCF7 cells were obtained from ATCC (HTB-22^TM), while MSCs were isolated using previously reported protocols [55,56]. The cells were cultured in DMEM/high glucose supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin at 37°C in a humidified atmosphere with 5% CO₂.

2.2.2. Cytotoxic assessment using MTT assay

The potential of eleven synthesized compounds (3a-k) to impede the proliferation of MCF7 cells was determined using the MTT experiment. The procedure involved seeding each well of a 96-well plate with 1 × 10⁵ MCF7 cells/mL and cultivating them in the growth media at 37°C for the whole night, where these cells were allowed to reach a monolayer confluency of 70% before undergoing any additional treatment. Subsequently, the cells were exposed to the test compounds for 24 hours at 37°C at a concentration of 100 µM. The cells were examined for toxicity indicators in the treated and control cells using an inverted microscope (Leica DMI3000B, Mannheim, Germany). After that, the MTT solution was added to the cells, which resulted in the conversion of MTT into formazan, the quantity of which corresponded to the total number of viable cells. The cells were then treated with dimethyl sulfoxide to remove the formazan crystal. Observations were noted at 570 nm using a microplate reader (Bio-Rad, Hercules, CA, USA) to assess the cytotoxic effect of the synthesized derivatives on the treated cells, and the percentage of cell inhibition was estimated. To further ascertain the safety of synthesized derivatives for utilization on living tissues. A dose-response curve of compound 3 h was employed to assess the half inhibitory dose (IC50) at a concentration ranging from 5 to 100 µM after a 24-hour incubation period in the MTT approach. The dose response curve of compound 3 h was compared with carboplatin (a commercially available anticancer drug) as a positive control.

To determine the cytotoxic effect of compound 3 h on normal cells, its dose response curve at concentrations ranging from 5 to 100 µM determined and IC50 against MSCs was calculated by subjecting dose response data to a linear regression ($\mathit{y}=\mathit{mx}+\mathit{b})$ in Microsoft Excel. The MSCs were isolated using previously reported methods [55,56]. The isolated MSCs were seeded at 1 × 10⁵ per well and cultured up to 70% confluency. The MSCs were then treated with different concentrations of compound 3h for 24 hours. Subsequently, MTT assay was performed to determine the cell viability.

2.2.3. Real-time quantitative PCR (RT-qPCR)

To identify the gene expression, MCF7 cells were cultured in 6 well plates at a seeding concentration of 5 × 10⁴ cells per well. The cells were divided into two groups, i.e., untreated (control) and treated (with compound 3 h, 16.5 µM). After 48 hours, the cells were trypsinized and the total mRNA was extracted using GeneJet RNA purification kit (Thermo Scientific) according to manufacturer’s instructions. Thereafter, complementary DNA (cDNA) was generated using revertAid cDNA synthesis kit (Thermo Scientific, MA, USA). The expression levels of the transcripts were determined with BlasTaq 2X qPCR MasterMix (G891, Applied Biological Materials) on StepOnePlus Real-time PCR system (Applied Biosystems). The ∆Ct values were calculated for each test gene normalized to human beta actin (β-actin) gene. The gene expression levels of the compound 3 h treated MCF7 cells were calculated as fold change relative to control MCF7 cells. The primer sequences are provided in (Supplementary Table S1). Statistical comparisons were performed by Wilcoxin test, and a p-value of less than 0.05 was considered statistically significant.

2.3. In silico study

2.3.1. Molecular descriptors

The prediction of physicochemical characteristics and drug likeliness of the synthesized derivatives were determined using ChemOffice 16.0 and Swiss ADME [57], as molecular weight (MW), hydrogen bond acceptors (HBA), number of rotatable bonds (RB), hydrogen bond donors (HBD), the logarithm of octanol/water partition coefficient (LogP), and topological polar surface area (TPSA). Furthermore, MW, RB, and TPSA elucidate the drug-likeness of a compound that can be determined using various MD criteria, including Lipinski’s rule of five (Ro5) [58,59], Veber, Egan, Muegge, and Ghose filter.

2.3.2. Calculation of bioactivity score (BAS)

Bioactivity score (BAS), a metric that evaluates a chemical entity’s overall drug-likeness, was calculated with Molinspiration Property Engine v2018.10 software. According to BAS rule, a higher reflected value indicates a greater probability of the compound being biologically active.

2.3.3. ADMET evaluation

Swiss ADME [57], pkCSM [60], ADMETlab 2.0 [61], PreADMET server (https://preadmet.webservice.bmdrc.org/) and ProTox-II [62] were used in this study to assess ADMET characteristics of the synthesized derivatives. Structural-based searches are employed using these databases to determine ADMET features.

2.3.4. Bioavailability radar (BAR)

Bioavailability radar (BAR) is valuable for evaluating the likelihood of the synthesized compound as a drug. Six physicochemical attributes (size, solubility, polarity, saturation, flexibility, and lipophilicity) comprise the pink region of the BAR plot. The oral bioavailability of the compound relies on its appearance within the pink region of the graph.

2.3.5. Brain or intestinal estimated (boiled-Egg)

The BOILED-Egg model assesses a compound’s propensity to pass through the BBB and the gastrointestinal system [63]. The drug-likeliness of the synthesized compounds was determined using this model. The two key factors that determine the bioavailability of a compound are lipophilicity (WLOGP) and TPSA. Using TPSA and WLOGP data, the Swiss ADME web server [57] creates a BOILED-Egg plot.

2.3.6. Molecular docking

2.3.6.1. Preparation of ligand

The 2D structure of chrysin derivatives and co-crystal ligand were drawn on the licensed software ChemOffice and saved in CDX format. The SMILE string of structure was generated by MarvinSketch 5.5 (Academic version).

2.3.6.2. Selection and preparation of protein

For the following study, the crystal structure of the target proteins Bax (PDB ID 6EB6) [64], Bcl-2 (PDB ID 6QGH) [65], p53 (PDB ID 5G4M) [66] and p21 (PDB ID 2CDZ) [67] was retrieved from the Protein Data Bank RCSB PDB Web server (www.rcsb.org/pdb/home/home.do) with a resolution of 2.02, 2.00, 1.38 and 2.30 Å, respectively. Using Discovery Studio Visualizer (DSV) 2021 atomic coordinates involve the water molecules, heteroatoms, and co-crystal ligand from the PDB file were removed, allowing the separation of protein structural inhibitor. SPDBV 4.2 was used to minimize the energy of protein.

2.3.6.3. Initialization and preparation of PDBQT files

The PDB format of the prepared protein molecule was imported in AutoDock Tool (1.5.7) workstation (http://autodock.scripps.edu/). To ensure proper tautomeric states and ionization of amino acid residues, polar H-atoms were incorporated into the target protein. The Kollman and Gasteiger charges were calculated, converting protein structure from PDB to PDBQT format. The conformational energies of the designed 2D analogs and standard were minimized and were converted from CDX to PDB format using Chem3D. Similarly, the PDB format of minimized ligand structures were loaded into the workspace where the torsion tree was defined by selecting the root and number of torsions, determining rotatable bonds, and converting into PDBQT format. This file format was then subjected to docking studies. Prior to this, a startup directory had been set to the specified folder.

2.3.6.4. Formation of the grid box

To perform further computations, it is essential to locate the protein’s active site for molecular docking. For Bax, the selected active site domain, including LEU 47, ARG 89, GLU 131, ARG 134, MET 137, GLY 138, and LEU 141. The grid’s dimensions were set to 102 × 72 × 42 xyz points, the center was marked × = 18.332, y = −19.484, and z = 1.472, with a spacing of 0.375Å [68]. For Bcl-2, the grid box enclosed the amino acid residues ALA 100, ASP 103, PHE 104, ARG 107, TYR 108, ASP 111, PHE 112, MET 115, VAL 133, GLU 136, LEU 137, ASN 143, TRP 144, GLY 145, ARG 146, VAL 148, ALA 149, GLU 152, PHE 153, VAL 156, PHE 198, LEU 201, TYR 202, GLY 203, ASN 204, ASN 205 and GLU 209. It was centered at x = -1.452, y = 3.586, z = 9.449, with dimensions of 68 × 88 × 92 xyz points and a grid spacing of 0.375 Å. For p53, the involved residues of chain A were PHE 109, LEU 145, TRP 146, VAL 147, THR 150, PRO 151, CYS 220, GLU 221, PRO 222, PRO 223, ASP 228, CYS 229 and THR 230. The grid box was centered at × = 91.384, y = 95.654, and z = −42.765, with dimensions of 56 × 60 × 54 and a spacing of 0.375 Å. For p21, the selected residues included ILE 327, GLY 328, GLU 329,VAL 335, ALA 348, VAL 379, MET 395, GLU 396, PHE 397, LEU 398, GLU 399, GLY 401, ALA 402, LEU 447, SER 457, with the grid dimensions and center of 64 × 52 × 70 and × = 44.100, y = 32.502, z = 20.666 along with a spacing of 0.375Å, respectively. These parameters were used to generate configuration files.

2.3.6.5. AutoDock Vina

The ligand-receptor interaction profile is analyzed using AutoDock Vina v1.1.2 [69]. This technique determines the interaction profile based on the minimum energy needed by systemically investigating the system’s degrees of freedom. Vina employs an empirical scoring approach and Broyden–Fletcher–Goldfarb–Shanno (BFGS) algorithm and performs better in docking estimation than AutoDock4 (AD4). It typically modifies the search space by the specifications and sets its mode number, energy range, and exhaustion to their default values. Although higher exhaustion values lengthen the search time, they also increase the precision of the findings.

2.3.6.6. Screening criteria

The results of the docking study were analyzed and visualized using the PyMOL molecular visualization tool and DSV 2021, which showed the size and location of the binding site, bond distances, and interactions of the docked ligand with the target.

2.3.6.7. Docking software validation

To validate the docking procedure a redocking and superimposition method was applied which involves the removal of co-crystal ligand (PDB ID 2CDZ, 5G4M, 6QGH) and re-docked it using AutoDock v4.2.6 into the active site. Throughout the process, identical protocol parameters including the grid specifications remained the same. Compared to the actual co-crystallized complex, the inhibitor must exhibit less deviation and bind exactly to the active domain. The redocked complex was then superimposed using PyMOL on the reference co-crystal complex to calculate root mean square deviation (RMSD).

2.3.7. Molecular dynamics (MD) simulation

The molecular dynamics (MD) simulation was performed by using GROningen MAchine for Chemical Simulations (GROMACS) version 2024.2 [70–72] to assess the stability of selected ligand-protein complex in docked state. On the basisof molecular docking and in-vitro study the compound 3h was selected for the MDS study. The four different complexes of compound 3 h with Bax, Bcl-2, p53, and p21 protein (PDB ID: 6EB6, 6QGH, 5G4M, 2CDZ) were considered for MD simulation study. The topology files of proteins were generated by GROMACS module “pdb2gmx” and the topology files of ligand were created by CGENFF server. Then, the protein-ligand complexes were immersed in a TIP3P water solution for solvation. The 3D cubic simulation boxes were created around selected protein-ligand complexes with the minimum distance of 1.0 nm between the complex and wall of box. Now, the solvent box was ready to solvate. In this stage the ions were added by “gmx genion” module to neutralize the complete system by introducing sodium and chloride ions. Furthermore, the energy minimization of system was performed using the steepest descent minimization algorithm with 50,000 of maximum steps with maximum force of < 10.0 kJ/mol and the ligand was restrained by using “gmx grompp.” Moreover, the system was equilibrated by two steps. In the first step, a NVT ensemble also called canonical ensemble with constant number of particles, volume, and maintained temperature. In the second step, NPT, also known as isothermal-isobaric, ensembles through constant pressure (1 atm), particles with maintained temperature. The NPT and NVT were employed for a 100-picosecond ensemble equilibration. The temperature was reduced to 310K by means of Berendsen thermostat technique. The particle mesh Ewald (PME) and van der Waals methods were used to calculate the electrostatic and van der Waals interactions, respectively, with a 1.2 nm cutoff and Fourier spacing [73]. Now, the well equilibrated and solvated system was subjected to the production phase for the period of 100 nanosecond with time step of 2 fs and after every 2 ps the structure coordinates were saved. The MD trajectories were analyzed in terms of RMSD, root mean square fluctuation (RMSF), radius of gyration (ROG), solvent-accessible surface area (SASA), and number of hydrogen bonds (HB) created between ligand and protein.

2.3.7.1. Binding free energy calculation

The free energy of protein with ligand is reflects the strength of their interaction. The value of energy represents the possibility of interaction. The gmx_MMPBSA module of GROMACS was used to generate the “mmpbsa.in” format. All frames of trajectory were used for analysis.

3. Results

3.1. Synthesis of chrysin derivatives

A series of 7-substituted phenyl pyrano derivatives of chrysin were synthesized following a reported protocol [74], as illustrated in Figure 1. Briefly, the one pot synthesis involves the reaction between chrysin (2), an ethanolic solution of various substituted benzaldehydes (1) and malononitrile in the presence of catalytic quantities of piperidine. This reaction yielded the corresponding derivatives of chrysin (3a – k) with yields ranging from 67% to 96%. The final product was isolated by evaporating the solvent and recrystallized using ethanol. The purity of the synthesized compounds was assessed using TLC and for the most potent compound (3 h), it was further confirmed by HPLC using mobile phase of 0.1% trifluoroacetic acid and acetonitrile (60:40). The substituted electron donating groups (EDGs) and electron withdrawing groups (EWGs) of benzaldehyde notably affect the reaction rate and product yield. Aryl aldehydes containing EWGs reacted more rapidly and produced higher yields than those with EDGs [74], as summarized in Table 1.

Figure 1.

Synthesis of chrysin derivatives via Michael-type addition reaction.

Table 1.

Synthesis of chrysin derivatives with varying substituents.

Compounds	Substituents (R)	Time (h)	Yield (%)
3a	H	8	80
3b	2-OH	9	74
3c	2-Cl	7	82
3d	2-formylpyridine	12	73
3e	2-Cl, 4-OCH3	7	87
3f	3-NO2	8	96
3g	3,4-(OCH3)2	7	75
3h	4-OH	8	70
3i	4-CH(CH3)2	15	72
3j	4-CH3	9	71
3k	4-N(CH3)2	11	67

In this reaction, substituted benzaldehyde and malononitrile undergo Knoevenagel condensation, eliminate a water molecule and produce an arylidine malononitrile as an intermediate. In the subsequent step, this intermediate experiences a Michael-type addition at the 8th position of chrysin, following that intramolecular cyclization, yielding the derivatives of chrysin.

Several techniques for spectroscopic analysis, including MS, FT-IR H¹ and C¹³ NMR spectroscopy, were employed to characterize the synthesized derivatives of chrysin. Among all the derivatives, compound 3 h was selected for detailed structural analysis using these spectroscopic methods: FT-IR, MS, H¹ NMR, and C¹³ NMR. The presence of stretching bands in the FT-IR spectrum of 3 h near 3520 cm⁻¹ and 2380 cm⁻¹, caused by – NH₂ and – CΞN, indicates that the structure of chrysin has been modified to the corresponding derivative. Compound (3 h) H¹ NMR spectrum revealed the generation of a pyran ring with a singlet peak at 4.98 ppm for the proton of the pyran ring and another singlet peak for the – NH₂ group’s 2 H at 4.31 ppm. The formation of the pyran ring is also supported by the C¹³ NMR spectrum, which shows a signal at 49.56 ppm corresponding to the pyran ring carbon. The molecular ion (M+) peak in the mass spectral analysis (EI+) of compound 3h appeared at m/z 424. As part of the fragmentation pattern, the loss of an NH₂ group resulted in a fragment peak at m/z 408, while the loss of the phenolic group gave rise to a base peak at m/z 331.

3.2. Exploring structure activity relationship (SAR) and their anticancer potential in vitro

The anticancer potential of synthesized derivatives (3a-k) against MCF7 cells was assessed using MTT test, with chrysin as the positive control. 100 μM was the sole concentration used for the initial screening (Supplementary Table S2). The common scaffold is a chromene core (8-amino-9-carbonitrile-10-pyrano-5-hydroxy-4-oxo-chromene) linked to two phenyl groups. One of these phenyl rings is substituted with various functional groups to explore their impact on anticancer properties. The 4-hydroxyl substituted phenyl derivative (3 h) showed remarkable anticancer activity against MCF7 cell line having IC50 value 16.5 μM which was greater than standard drug carboplatin (IC50 53.3 μM) and chrysin (IC50 > 100 μM) while the hydroxyl group at position 2 (3b) show no significant activity. The 4-dimethylamino substituted phenyl derivative (3k) showed significantly higher activity (55.83% inhibition) compared to chrysin. Replacing the dimethylamino group with an isopropyl group (3i) completely eliminated activity. However, incorporating a methyl group at the 4-position (3j) or methoxy groups at both the 3 and 4 positions (3 g) resulted in moderate activity. The 3-nitro substituted phenyl derivative (3f) exhibited activity comparable to chrysin. The 2-chloro substituted phenyl derivative (3c) showed similar activity to chrysin (30.0% inhibition), but adding a methoxy group (3e) decreased its effectiveness, suggesting that increased hydrophobicity on the phenyl ring can negatively impact activity. The unsubstituted phenyl derivative (3a) showed no activity, while replacing the phenyl ring with a pyridine ring (3d) enhanced activity, indicating a balance between hydrophobicity and hydrophilicity is needed for optimal activity.

The cytotoxic effect of compound 3 h on normal cells against human MSCs was determined. The results showed that the compound 3 h was less cytotoxic on MSCs as compared to MCF7 cells (Figure 2(D)). At a 100 μM concentration it caused an inhibition of 52.58% in MSCs as compared to 95.11% in MCF7 cells. In addition, it displayed a much higher IC50 value of 77.95 μM against MSCs. These findings indicate that at an IC50 concentration of 16.5 μM against MCF7 cells, the compound 3 h will cause a much lower cytotoxic effect on normal cells.

Figure 2.

Dose response and gene expression analysis. (A) Anticancer activity of all compounds against MCF7 cells at 100 µM concentration. (B) Dose response curve of compound 3 h and carboplatin to determine its IC50 concentration. (C) Gene expression levels of MCF7 cells treated with compound 3 h (16.5 µM) relative to untreated MCF7 cells (n = 3, p > 0.05 = non-significant) (D) Dose response curve of compound 3 h to determine its safety profile in MSCs.

3.3. RT-qPCR analysis

The impact of compound 3 h on the expression of pro- and anti-apoptotic proteins was evaluated using RT-qPCR analysis. The expression levels of the tumor suppressor genes p21 and p53, the pro-apoptotic gene Bax and the anti-apoptotic gene Bcl-2 were assessed in each experimental group. As depicted in Figure 2(C), the expression levels of Bax, Bcl-2 and p53 were significantly higher in the cells treated with compound 3 h compared to the untreated group. In contrast, treatment of MCF7 cells with compound 3 h reduces p21 expression, suggesting that p21 plays a function in cell cycle regulation.

3.4. Molecular descriptors

Tables 2 and 3 display the drug-like molecular characteristics of the synthesized derivatives which were forecast using freely accessible web applications. These derivatives satisfy Lipinski’s Rule of Five requirements (MW ≤ 500, HBD ≤ 5, HBA ≤ 10 and log p ≤ 5) and Ghose Filters (MW 160–480, #atoms 20–70, PSA ≤ 140, molar refractivity 40–130 and log P −0.4–5.6). One compound 3f, violates Veber rule (#RB ≤ 10 and polar surface area ≤ 140) and Egan rule (TPSA ≤131.6Å or log p ≤ 5.88), the predicted polar surface area was found to be 155.30. In addition to this, five compounds 3c, 3e, 3f, 3i and 3j out of eleven breach Muegge rule (200 ≤ MW ≤ 500, HBA ≤ 10, HBD ≤ 5, # RB ≤ 15, # rings ≤ 7, # C ≥ 4, # Hetro atom ≥ 1, TPSA < 150, −2 ≤ ClogP ≤ 5). Among them, compound 3c, 3e and 3i log p values were > 5, whereas 3f deviates from the TPSA criteria of the Muegge rule. The structural features of the derivatives also comply with the CMC 50 Like rule and the BBB rule. Based on these projected values, it was possible to infer that the improved oral bioavailability and distribution of derivatives were related to their structural characteristics and physicochemical qualities.

Table 2.

Rules of drug likeness.

Rules	Compounds
	3a	3b	3c	3d	3e	3f	3g	3h	3i	3j	3k
Lipinski	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes
Veber	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes
Egan	Yes	Yes	Yes	Yes	Yes	No	Yes	Yes	Yes	Yes	Yes
Muegge	Yes	Yes	No	Yes	No	No	Yes	Yes	No	No	Yes
Ghose	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes	Yes

Table 3.

Drug-like molecular and physicochemical properties of synthesized derivatives.

	Compounds
	3a	3b	3c	3d	3e	3f	3g	3h	3i	3j	3k
Mol. Formula	C25H16N2O4	C25H16N2O5	C25H15ClN2O4	C24H15N3O4	C26H17ClN2O5	C25H15N3O6	C27H20N2O6	C25H16N2O5	C28H22N2O4	C26H18N2O4	C27H21N3O4
Mol. Mass (g/mol)	408.41	424.40	442.85	409.39	472.88	453.40	468.46	424.40	450.49	422.43	451.47
tAtoms	47	48	47	46	51	49	55	48	56	50	55
Heavy Atoms	31	32	32	31	34	34	35	32	34	32	34
nRing	5	5	5	5	5	5	5	5	5	5	5
nHD	2	3	2	2	2	2	2	3	3	2	2
nHA	5	6	5	6	6	7	7	6	5	5	5
nRot	2	2	2	2	3	3	4	2	3	2	3
TPSA (Å2)	109.48	129.71	109.48	122.37	118.71	155.30	127.94	127.71	109.48	109.48	112.72
Molar refractivity (cm3/mol)	115.37	117.39	120.38	113.17	126.87	124.19	128.35	117.39	129.95	120.34	129.58
Fraction of SP3 carbon	0.04	0.04	0.04	0.04	0.06	0.04	0.11	0.04	0.14	0.08	0.11
log P	4.38	4.09	5.04	3.78	5.05	4.29	4.40	4.09	5.51	4.69	4.45
pKa	20	20	20	20	20	20	20	20	20	20	20

3.5. Bioactivity score (BAS) analysis

Bayesian statistics [75] were employed to project the outcomes of the bioactivity score as presented in Supplementary Table S3. The analysis revealed six major pharmacological classes: ion-channel modulators, nuclear receptor, G-protein-coupled receptor ligands, and enzyme inhibitors, all contrasted by their derivatives. The computed results indicate that the synthesized derivatives demonstrate considerable bioactivity.

3.6. ADMET profile

The pharmacokinetic and toxicological parameters of the proposed derivatives are shown in Supplementary Table S4. These were analyzed using open-source web servers to obtain predicted values of their absorption parameters. As the Table shows, these compounds have poor buffer and water solubility. The human intestine was the site of their excellent absorption, ranging from 93.07% to 100% human intestinal absorption rate (the in-silico permeability at Caco-2 was ≤ 0.75 log Papp in 10 cm/s). The skin permeation was extremely low, −2.735 log Kp cm/h. These compounds are a substrate of the ATP binding cassette (ABC) transporter that acts as a bio-defense mechanism against toxins and foreign substances. Moreover, they inhibited P-glycoproteins.

According to Supplementary Table S5, the volume of distribution for proposed derivatives in humans was less than 0.055 log L/kg, indicating a well-dispersed substance in tissues. The CNS (log PS) and BBB (log BB) permeability values were notably low, implying that these substances do not permeate the brain and the central nervous system. The synthesized derivatives were evaluated against a range of enzyme substrates and inhibitors as displayed in the Supplementary Table S6, including CYP2D6, CYP3A4, CYP1A2, CYP2C19, CYP3A4, CYP2D6 and CYP2C9. The compounds were inhibitors of CYP2C9, CYP2C19, CYP3A4 and substrate exclusively for CYP3A4; a small number of derivatives also inhibited CYP1A2. The expected excretion parameters of derivatives (Supplementary Table S7) indicate that each compound has a high overall renal clearance rate and does not interact with the renal OCT2 as its substrate.

According to the GHS (globally harmonized system), all synthesized derivatives belong to toxicity class 4, indicating their likelihood of being less toxic (LD50 300–2000 mg/kg) [76]. The rats’ acute (mg/kg) and chronic oral toxicity dose (log mg/kg/day) were determined using pkCSM. These derivatives do not cause considerable acute oral toxicity to daphnia, flathead minnow fish (log mM) and T. pyriformis (log µg/L). They are hepatotoxic. However, these are less likely to be skin-sensitizing, carcinogenic, or cardiotoxic agents. Additionally, these derivatives tend to inhibit hERG-II but were not inhibitors of hERG-I. Due to their inability to bind with common toxicity targets, such as the kappa-type opioid receptor, D3 dopamine receptor, nuclear receptor subfamily 1 and 3, histamine H1 receptor, beta-2 adrenergic receptor androgen receptor, adenosine receptor A2a, estrogen receptor, and glucocorticoid receptor, as well as the corticotropin-releasing factor receptor, they do not cause serious side effects. The toxicological profile of the derivatives is present in Supplementary Table S8.

3.7. Bioactivity radar (BAR) and BOILED–Egg assessment

The BAR model offers a compendious visual depiction of a drug’s attributes based on its molecular features. The range of characteristics, such as size, polarity, flexibility, solubility, and lipophilicity, are ideal for oral bioavailability, as shown by the pink region in the BAR. Based on the BAR prediction, the synthesized derivatives were in the desired pink region. Additionally, the BOILED-Egg model evaluates the absorption and permeability of substances throughout the gastrointestinal and BBB [77]. According to this model, the yellow yolk denotes the capability to pass across the BBB, while the white portion reflects the probability of gastrointestinal absorption. The findings reveal that none of the compounds can pass across the BBB and only a small number of them are absorbed effectively by the gastrointestinal tract (Supplementary Figure 45). These results collectively suggest that the synthesized derivatives are likely to be a potential biologically active.

3.8. Molecular docking

AutoDock Tools were utilized to bind the proposed chemical structure of synthesized derivatives to the active site of the Bax, Bcl-2, p53, and p21. Binding affinities of the derivatives with the target proteins are summarized in Supplementary Table 9S. Binding interactions of compound (3 h), an intriguing candidate with notable anticancer activity, were analyzed using DSV 2021. Table 4 presents binding energy of the compound 3 h with the target proteins. Figure 3 illustrates the best compound orientation and binding affinity at an active site foreseen from the docking complex. The validity of the docking method was confirmed using the RMSD that was calculated between the co-crystallized ligand and the redocked pose for the Bcl-2, p53 and p21 protein was < 2 Å, falling within an acceptable range [78].

Table 4.

Binding energy and type of interactions computed by molecular docking between the compound (3 h) and the target proteins.

Target Protein	Binding Energy (kcal/mol)	Hydrogen Bonding (Å)	π-Interactions (Å)	Van der Waals
Bax	−9.2	–	ASP 84 (6.44), (5.63), THR 127 (5.17), LYS 123 (5.47), (4.74), (4.83), PRO 88 (5.93), ALA 124 (5.46)	VAL 83, THR 85, LEU 120, CYS 126, VAL 129, LEU 132
Bcl-2	−8.9	ARG 107 (3.46), ASN 143 (4.74)	ARG 107 (5.40), ALA 100 (5.54), LEU 201 (6.22), PHE 198 (6.11)	ASP 103, PHE 104, TYR 108, TRP 144, GLY 145, VAL 148, TYR 202, ALA 206, GLU 209, SER 210
p53	−8.0	SER 261 (5.54), ARG 267 (5.91)	ASP 208 (7.25), THR 256 (4.47), PRO 98 (5.74), MET 160 (5.99), ILE 254 (5.99), LEU 264 (6.05)	SER 99, ARG 158, LEU 206, THR 211, ARG 213, GLU 258, GLY 262, ASN 263
p21	−9.4	GLU 329 (4.88), (4.92), GLY 401 (3.91)	ILE 327 (5.11), VAL 335 (5.45), LEU 447 (5.49), ALA 348 (6.60)	GLY 328, GLY 330, PHE 397, LEU 398, GLY 401, ALA 402, ASP 405

Figure 3.

Visualization of binding interactions of compound 3 h with the target proteins using discovery studio visualizer 2021. (A) Interaction of compound 3 h with the binding residues of Bax (PDB ID 6EB6). (B) Interaction of compound 3 h with the binding residues of Bcl-2 (PDB ID 6QGH). (C) Interaction of compound 3 h with the binding residues of p53 (PDB ID 5G4M). (D) Interaction of compound 3 h with the binding residues of p21 (PDB ID 2CDZ).

The binding mode of the compound (3 h) at the Bax active site yielded a docking score of −9.2 kcal/mol. According to the docked complex, the active compound’s chromen core exhibited two π-anion interactions with ASP 84, one π-sigma bond with THR 127 and a π-stacked bond with LYS 123. Similarly, the phenyl rings of the compound formed four π-alkyl bonds with Pro 88, Lys 123, Ala 124 and six van der Waals interactions with VAL 83, THR 85, LEU 120, CYS 126, VAL 129, and LEU 132 residues.

The docking of compound (3 h) with Bcl-2 demonstrates a binding energy of −8.9 kcal/mol which results in the formation of two hydrogen bonds between the nitrile moiety of the pyran ring and the hydroxyl moiety of the chromen core with ASN 143 and ARG 107. Likewise, ALA 100, PHE 198 and LEU 201 established three hydrophobic bonds with the phenyl ring, whereas the chromen core made a π-cation bond with ARG 107 residue. Compound 3 h also exhibit ten van der Waals interactions with ASP 103, PHE 104, TYR 108, TRP 144, GLY 145, VAL 148, TYR 202, ALA 206, GLU 209, and SER 210.

The binding interaction of compound (3 h) with p53 reveals that two hydrogen bonds, one π-sigma bond, one π-anion bond and four π-alkyl bonds with the residues, as demonstrated in Figure 3, were crucial, followed by a docking score of −8.0 kcal/mol. SER 261 and ARG 267 were involved in the hydrogen bonds between the amino moiety of the pyran ring and the carbonyl moiety of the chromen core. Additionally, phenyl rings formed one π-anion bond with ASP 208, one π-sigma bond with THR 256 and four π-alkyl bonds with PRO 98, MET 160, ILE 254 and LEU 264. Similarly, eight residues SER 99, ARG 158, LEU 206, THR 211, ARG 213, GLU 258, GLY 262, and ASN 263 were involved in der Waals interactions.

The docking analysis of compound (3 h) with p21 revealed that the docked complex displayed a docking score of −9.4 kcal/mole. Similar to the interaction with Bcl-2, compound (3 h) forms two hydrogen bonds between the hydroxyl group of the chromen core with GLU 329 and one hydrogen bond between the nitrile moiety of the pyran ring with GLY 401. Further, phenyl rings interact hydrophobically to form three alkyl bonds and one π-alkyl bond with VAL 335, ALA 348, LEU 447, and ILE 327 residue. Besides this, seven van der Waals interactions with GLY 328, GLY 330, PHE 397, LEU 398, GLY 401, ALA 402, and ASP 405 were also a part of 3 h-protein complex.

3.9. Molecular dynamic simulation

The MD simulation was performed for compound 3 h with Bax, Bcl-2, p53 and p21 proteins (PDB ID: 6EB6, 6QGH, 5G4M, 2CDZ). Figure 4 presents the conformational changes at different dynamic allosteric states, and it was observed that the compound 3 h is stable in active pocket of selected protein at 0, 20, 40, 60, 80, and 100ns. The stability of protein-ligand complexes was analyzed in terms of RMSD. The average RMSD values of protein backbone atoms in each complex were 0.16, 0.18, 0.21, and 0.72 nm for 3 h-6EB6, 3 h-5G4M, 3 h-2CDZ, and 3 h-6QGH respectively (Figure 5(A)). During MD simulation, the RMSD values of the studied complexes exhibited consistently low.

Figure 4.

Conformational changes of compound 3 h in the active pockets of Bax (PDB ID 6EB6), p53 (PDB ID 5G4M), p2 1 (PDB ID 2CDZ), and Bcl-2 (PDB ID 6QGH) proteins at dynamic allosteric states observed at 0, 20, 40, 60, 80, and 100 ns during molecular dynamics (MD) simulations.

Figure 5.

Comparative plots of (A) root mean square deviations (RMSD), (B) solvent accessible surface area (SASA), (C) radius of gyration (ROG), and (D) number of hydrogen bonds for Bax (PDB ID 6EB6), p53 (PDB ID 5G4M), p21 (PDB ID 2CDZ), and Bcl-2 (PDB ID 6QGH) proteins in complex with compound 3 h.

3.9.1. Binding free energy calculation

The binding free energy was calculated by using MM-PBSA method implemented in the gmx_MMPBSA tools [79]. The negative values for Gibbs free energy indicate that the ligand binding process is exothermic and it is thermodynamically stable. The graphical representation was visualized by gmx_MMPBSA_ana tool and the components of energy were calculated by MM/GBSA method. Supplementary Figure 46 represented the energy values (ΔGGAS, ΔGSOLV, and ΔGTOTAL) of complexed 3 h with selected proteins. Total binding free energy were −17.25, −19.84, −16.08, and −20.59 Kcal/mol of 3 h-6EB6, 3 h-5G4M, 3 h-2CDZ, and 3 h-6QGH respectively.

4. Discussion

Chrysin derivatives, with a core chromene structure, are being investigated for their anticancer potential, particularly those with modifications on the phenyl rings. Specifically, the 4-hydroxy and 4-dimethylamino substituted phenyl derivatives (compounds 3 h and 3k, respectively) show promising anticancer activity, with compound 3 h outperforming the parent chrysin and standard carboplatin.

Molecular docking studies facilitate the drug design process by predicting the affinity and binding pose of compounds with the target [80]. The results of our work, which involved docking the potent chrysin derivative with Bax, Bcl-2, p53 and p21 protein active site, demonstrated that 4-hydroxy substituted phenyl derivative (3h) possess excellent binding affinities against all the targets. Generally, a negative score depicts that the compound can bind to the target’s active site with good affinity. Within the binding pocket, 3 h participated in various bond formation processes, such as π interactions, hydrogen bonds and van der Waals forces. These interactions are crucial for its pharmacological effects and binding to the target [81].

Furthermore, MD simulation was used to assess the behavior and stability of a compound, and the results were deemed satisfactory. This approach involves simulating the interaction between a Bax, Bcl-2, p53, and p21 protein and compound 3 h to predict how they might behave in a biological environment, specifically looking at thermodynamic stability in active site of protein at 0, 20, 40, 60, 80, and 100ns as (Figure 4). The graph of RMSD specifies the stable curves for all selected ligand-protein complexes during the simulation time that was 100ns and also indicates that studies complex was stable during simulation period. The RMSF values of each residue in each complex were calculated and depicted in Supplementary Figure 47. In 3 h-5G4M complex the fluctuation was seen in the zone of 176–194, 201–203, 219–231, and 262 (Figure 47A). Similarly, in 3 h-2CDZ complex the fluctuation was observed in 3 different regions 350–364, 462–485, and 514–546. It was indicated that these three-regions involved in protein ligand interactions (Figure 47B). In 3 h-6QGH complex only single fluctuation was present in the region of 205–220 (Figure 47C). Finally, the 3 h-6EB6 complex produced fluctuation in section of 34–59, 81–91, 102, and 191 (Figure 47D). Additionally, the complexes were also evaluated by SASA. The value SASA indicates the degree of hydrophobicity of protein. The large values correspond the increased protein content expose to the water and less temporal variability. The average values were 94.89, 108.75, 149.36 and 98.46 nm for 3 h-6EB6, 3 h-5G4M, 3 h-2CDZ, and 3 h-6QGH complex respectively. The results were compiled in Figure 5B, which indicated that the proteins were stable during simulation study. Furthermore, the complexes were also evaluated by ROG. The radius of gyration is used to measure the compactness of protein. The low value of ROG indicated that the studied protein was compact and folded and the high values indicated that the structure of protein was expended and unfolded. The average values of ROG were 1.57, 1.68, 1.99 and 1.65 nm for the complexes of 3 h-6EB6, 3 h-5G4M, 3 h-2CDZ, and 3 h-6QGH respectively. The results of ROG for 100 ns were presented in Figure 5C which revealed that the trajectory of the radius of gyration indicates the absence of anomalous behavior. The formation of hydrogen bond is a capacity of ligand to interact protein at active site. The results of simulations confirmed that the HB interaction between the ligand and protein showed a state of sustained and enduring stability (Figure 5D). The result disclosed that the compound 3 h was stable during the period of 100 ns. According to the above observation, the binding energy of the complex of 3 h-6QGH with receptor is the least and is more stable than complexes.

To validate the results of molecular docking and MD simulations, experimental validation using RT-qPCR to assess gene expression levels of Bax, Bcl-2, p53, and p21 in cancer cell line was investigated according to reported protocol. This allows for a direct comparison between the computational predictions and the actual biological response in a cellular or tissue system. The current study utilizes β-actin as a reference gene [82,83] for RT-qPCR investigations in MCF7 cells and demonstrated a significant increase in Bax, Bcl-2, and p53 mRNA expression levels. When activated in response to oxidative stress or DNA damage, p53 acts as a transcription factor that regulates genes involved in both cell cycle arrest and apoptosis [84,85]. In our case, Bax and Bcl-2 were both upregulated, but importantly, Bax showed a substantially higher fold increase than Bcl-2, resulting in an elevated Bax/Bcl-2 ratio. This ratio is a critical determinant of mitochondrial outer membrane permeabilization and commitment to apoptosis [86]. While Bax – Bax homodimers promote apoptosis, Bcl-2–Bax heterodimers inhibit it [87,88]. Thus, despite the observed upregulation of Bcl-2, which may reflect a cellular attempt to resist death – the predominance of Bax supports the conclusion that p53-induced apoptosis was successfully triggered. Additionally, although p21 expression was downregulated, this does not contradict apoptotic induction. While p21 is a key mediator of p53-dependent cell cycle arrest via CDK inhibition [15,89], its downregulation in this context may indicate that the cells bypassed arrest and instead committed to apoptosis, possibly due to the dominance of pro-apoptotic signals such as elevated Bax and p53 levels. The high cytotoxicity observed in our study further supports the conclusion that the pro-apoptotic arm of the p53 pathway (via Bax) outweighed the cell cycle arrest pathway (via p21) in determining cell fate.

5. Conclusion

Eleven chrysin-based molecules have been synthesized containing common scaffold (8-amino-9-carbonitrile-10-pyrano-5-hydroxy-4-oxo-chromene) the variations were made on the 10th position of phenyl ring. 3 h demonstrates potent cytotoxic activity against MCF7 breast cancer cells, exceeding that of carboplatin. It shows a strong binding affinity to target proteins (Bax, Bcl-2, p53, and p21) and stable interactions, as confirmed by molecular dynamics simulations. However, future studies with a broader range of cancer cells, assays including Western blots, and ultimately animal-based experiments will be required to conclusively ascertain the drug’s anticancer potential.

Funding Statement

This manuscript was funded by Higher Education Commission of Pakistan/National Research Program for Universities [HEC/NRPU # 15377].

Article highlights

• The derivatives of naturally occurring chrysin were synthesized by employing Michael-type addition reaction.

• In silico and in vitro models were used to assess the anticancer properties of the compounds.

• Among the synthesized derivatives, 4-hydroxy substituted phenyl derivative (3h) was identified as the most potent anti-breast cancer agent.

• 3h have a more profound effect than carboplatin along with its outstanding safety profile.

• Molecular docking and molecular dynamic simulation confirmed strong binding and complex stability.

• ADMET characteristics indicate that it has the potential to be developed into pharmaceutical.

Declaration of competing interest

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties

Writing Assistance

No writing assistance was utilized in the production of this manuscript

Reviewer disclosures

Peer reviewers on this manuscript have no relevant financial or other relationships to disclose.

Author’s contribution

Mahwish Akhtar wrote the project (HEC/NRPU # 15377) for grant, trained the student for all type of in silico work. Asia Naz Awan designed the experiments for synthetic work and biological activities and trained the student. Omair Anwar Mohiuddin had performed and wrote the biological activity part with Kainat Ahmed. Sabahat Abdullah performed all experiments and prepared first draft of article.

Declaration of generative AI and AI-assisted technologies in the writing process

To improve clarity and correct grammar, the author used ChatGPT while preparing this article. The author took full responsibility for the publication’s content after using this tool, reviewing and editing it as necessary.

Supplemental data

Supplemental data for this article can be accessed online at https://doi.org/10.1080/17568919.2025.2559569.