Discovery of new thiadiazole-based VEGFR-2 inhibitors: design, synthesis, cytotoxicity, and apoptosis induction

ABSTRACT

Background

Vascular endothelial growth factor receptor-2 (VEGFR-2) is a validated target in cancer therapy. However, approved inhibitors like sorafenib are often limited by off-target toxicity and resistance. This study aimed to develop novel thiadiazole-based VEGFR-2 inhibitors with improved selectivity and safer profiles.

Methods

A series of 2,3-dihydro-1,3,4-thiadiazole derivatives was rationally designed, synthesized, and structurally characterized. In vitro cytotoxicity was assessed against MCF-7, HepG-2, HCT-116, and normal WI-38 cells. Mechanistic assays included flow cytometry for cell cycle and apoptosis, and ELISA for caspase-3, Bcl-2, and Bax expression. Molecular docking, 200-ns molecular dynamics (MD) simulations, density functional theory (DFT) calculations, and in silico toxicity profiling supported experimental findings.

Results

Compound 11a exhibited the most potent and selective activity (IC₅₀: 9.49 µM for MCF-7, 12.89 µM for HepG-2; SI > 3). It induced >70% apoptosis and dual-phase (S and G2/M) cell cycle arrest. VEGFR-2 inhibition (IC₅₀ = 0.055 µM) was comparable to sorafenib. Computational studies confirmed stable binding at VEGFR-2 active sites.

Conclusion

Compound 11a is a promising thiadiazole-based candidate with notable in vitro potency, selectivity, and mechanistic activity, supporting its potential for further pharmacokinetics/toxicity evaluation and structural refinement as a VEGFR-2–targeted agent.

Graphical Abstract

1. Introduction

Cancer is still one of the leading causes of death around the world and continues to place a heavy burden on healthcare systems. Its widespread impact touches millions of lives and presents serious public health challenges, especially because diagnosing, treating, and managing the disease over time is often complex and demanding [1,2].

Vascular endothelial growth factor receptor-2 (VEGFR-2) plays a crucial role in angiogenesis and cancer progression, positioning it as an important molecular target in oncology. It is frequently overexpressed in breast cancer, where its elevated levels are strongly associated with poor clinical outcomes, increased blood vessel formation, and a greater risk of metastasis [3]. As a result, targeting VEGFR-2 has emerged as an effective therapeutic strategy to inhibit tumor angiogenesis and curb the progression of cancer [4]. While clinically approved VEGFR inhibitors have shown therapeutic benefits [5–7], their clinical utility is often constrained by notable side effects and limitations include drug resistance [8,9], undesirable off-target effects [10–12], and suboptimal pharmacokinetic properties [13], underscoring the urgent need for safer and more efficient treatment options.

In light of these challenges, thiadiazole-based scaffolds have attracted considerable attention for their broad pharmacological potential, particularly their anti-inflammatory [14,15], kinase-inhibitory [16,17] and anticancer [18–25] activities.

The integration of in silico and in vitro approaches offers a powerful and efficient strategy for modern drug discovery efforts [26,27]. This dual approach leverages the predictive strength of computational modeling (which can simulate molecular interactions, binding affinities, and pharmacokinetic profiles) alongside the biological relevance and mechanistic validation provided by experimental assays [28,29]. When used in tandem, these methods not only streamline the identification and optimization of potential therapeutic candidates but also improve the accuracy and reproducibility of results while minimizing the need for time-consuming and costly in vivo studies. Our research team has successfully utilized this integrated approach targeting SARS-CoV-2 [30] as well as microbial virulence factors [31], and cancer [32].

Building on our team’s earlier research, we present the design, synthesis, and biological evaluation of a series of novel thiadiazole scaffolds as potential anticancer VEGFR-2 inhibitors. In vitro assays were carried out to evaluate their cytotoxic effects, influence on the cell cycle, and ability to induce apoptosis. Furthermore, a range of comprehensive computational studies, including molecular docking, 200 ns MD simulations, interaction mapping, PCA-T analyses, in silico ADMET (absorption, distribution, metabolism, excretion, and toxicity) and toxicity studies. Finally, DFT calculations were employed to explore the binding mechanism, conformational stability, and electronic properties of the most promising compound.

Thiadiazole scaffolds were chosen due to their proven bioisosteric properties, favorable electronic features, and prior success in kinase inhibition. Specifically, the 1,3,4-thiadiazole ring can mimic key interactions within the VEGFR-2 ATP-binding site, including hydrogen bonding with hinge region residues [33]. However, few thiadiazole-based VEGFR-2 inhibitors have achieved an optimal balance of potency, selectivity, and low toxicity [34]. This study addresses this gap by designing novel derivatives incorporating key pharmacophoric features to enhance biological activity and safety.

From a synthetic standpoint, ethanol was selected as a green, cost-effective solvent that provided good solubility and yields under mild reflux conditions. Compared to polar aprotic solvents like DMF (dimethylformamide) or DMSO (dimethyl sulfoxide), often used in similar syntheses, ethanol minimized side reactions and simplified purification. Triethylamine was preferred as a mild base to facilitate cyclization without promoting degradation of sensitive intermediates [35].

1.1. Rationale of molecular design

A comprehensive literature review has established that numerous VEGFR-2 inhibitors have been successfully developed and approved as powerful anticancer agents. Notable examples include sorafenib (1) [36], tivozanib (2) [37] (Figure 1(A)). Clinical studies have demonstrated that these inhibitors exhibit broad-spectrum anticancer efficacy while being less susceptible to drug resistance. However, their clinical utility is often hampered by adverse effects such as fatigue, anemia, diarrhea, hypertension, and hand-foot skin reactions [38]. Consequently, there is an urgent and unmet need for next-generation VEGFR-2 inhibitors that offer enhanced therapeutic potency while minimizing toxicity [39].

Figure 1.

Rationale of molecular design. A) Basic pharmacophoric features of some reported VEGFR-2 inhibitors. B) The basic structural requirements for sorafenib (1) as reported VEGFR-2 inhibitor. C) Summary of the possible modifications.

An in-depth analysis of the structure–activity relationships (SAR) of reported VEGFR-2 inhibitors has revealed that their pharmacophoric architecture is defined by four key structural features (Figure 1(B)) [40–42]. (a) A hetero-aromatic ring system that interacts with the hinge region of the ATP-binding site, playing a pivotal role in molecular recognition [40]. (b) A central aromatic linker bridging the hinge region and the DFG domain is essential for optimal spatial orientation and target engagement [43]. (c) A pharmacophoric moiety incorporating at least one hydrogen bond donor (HBD) and one hydrogen bond acceptor (HBA), which is vital for receptor binding. Notably, the critical amino acid residues Asp1044 and Glu883 within the DFG domain have been identified as key interaction sites for this moiety [44]. (d) A terminal lipophilic group that occupies the allosteric lipophilic pocket, facilitating strong hydrophobic interactions that contribute to binding affinity and selectivity [45].

Building upon these established findings and as a natural extension of our ongoing research dedicated to designing novel anticancer agents, we embarked on the rational design and synthesis of an innovative series of 2,3-dihydro-1,3,4-thiadiazole derivatives. These new compounds were meticulously engineered to incorporate the fundamental pharmacophoric features essential for potent VEGFR-2 inhibition. Our molecular design strategy is centered on strategic modifications at four critical structural positions (Figure 1(C)). i) Hetero-aromatic core: We introduced diverse 3-phenyl-2,3-dihydro-1,3,4-thiadiazole moieties to enhance steric interactions and improve binding affinity with the ATP hinge region. ii) Central aromatic linker: The (1-phenylethylidene)hydrazine moiety was chosen to enhance molecular flexibility, optimizing the spatial arrangement for receptor interaction. iii) Pharmacophoric moiety (HBD/HBA): The amid functional group was strategically incorporated to strengthen hydrogen bonding with key residues, thereby improving target specificity. iv) Terminal lipophilic group: A para-tolyl moiety was selected to maximize hydrophobic interactions within the allosteric lipophilic pocket, thereby increasing potency and selectivity.

The structural variations introduced in this study enabled a focused SAR analysis and led to the identification of novel thiadiazole-based VEGFR-2 inhibitors with improved selectivity and anticancer activity. The designed compounds incorporate essential pharmacophoric features (Figure 1(C)) and provide new insights into how electronic and spatial changes influence bioactivity. Unlike prior reports, the substitution patterns and scaffold design used here are distinct and systematically evaluated. Supported by computational, mechanistic, and ADMET analyses, this work advances the development of safer, more selective VEGFR-2 inhibitors. Compound 11a, in particular, shows promise for further structural refinement and formulation enhancement as a targeted anticancer agent.

2. Experimental

2.1. Chemistry

Conventional chemical techniques were utilized for the synthesis and characterization of the compounds. The Supplementary Materials provide comprehensive experimental protocols. Compounds 7, 8, 9, 10a-c, and 12a-c were synthesized following established protocols [46–50].

2.1.1. General procedure for the synthesis of the final compounds 11a-c

The hydrazonoyl derivatives 10a-c underwent a reaction with an equal amount of compound 9 in 20 mL of absolute ethanol, employing 1 equivalent of triethylamine as a catalyst. All ingredients were heated under reflux for 6 h. Following the filtration and washing of the precipitates with hot ethanol, the final thiadiazole derivatives 11a-c were successfully obtained.

2.1.1.1. N-(4-((E)-1-(((Z)-5-Acetyl-3-(2-chlorophenyl)-1,3,4-thiadiazol-2(3 H)-ylidene) hydrazono)ethyl)phenyl)-4-methylbenzamide 11a

Yellow powder (yield, 75%); m.p. = 190–192°C. FT-IR (v max, cm⁻¹): 3339, 3061, 3029, 2966, 2919, 1671, 1657, 1602; ¹H NMR (400 MHz, DMSO-d₆) δ_H (ppm) = 10.32 (s, 1H, NH), 7.97–7.76 (m, 9H, Ar-H), 7.64–7.59 (m, 1H, Ar-H), 7.35 (d, J = 7.5 Hz, 2H, Ar-H), 2.53 (s, 3H, CH₃), 2.39 (s, 3H, CH₃), 2.21 (s, 3H, CH₃); ¹³C NMR (101 MHz, DMSO-d₆) δ_C (ppm) = 189.59, 165.46, 159.86, 141.77, 140.85, 135.15, 132.24, 131.87, 131.71, 131.08, 130.52, 130.15, 128.94, 128.51, 127.77, 127.00, 126.85, 119.80, 24.98, 21.03, 14.72; Mass (m/z): 504 (M⁺, 10.93%), 237 (100%, base peak); Anal. Calcd. for C₂₆H₂₂ClN₅O₂S (504.01): C, 61.96; H, 4.40; N, 13.90. Found: C, 62.11; H, 4.56; N, 13.77%.

2.1.1.2. N-(4-((E)-1-(((Z)-5-Acetyl-3-(m-tolyl)-1,3,4-thiadiazol-2(3 H)-ylidene)hydrazono)ethyl) phenyl)-4-methylbenzamide 11b

Yellow powder (yield, 77%); m.p. = 198–200°C. FT-IR (v max, cm⁻¹): 3340, 3060, 3029, 2966, 2919, 2863, 1672, 1657, 1601; ¹H NMR (400 MHz, DMSO-d₆) δ_H (ppm) = 10.34 (s, 1H, NH), 7.93–7.85 (m, 8H, Ar-H), 7.47 (t, J = 7.8 Hz, 1H, Ar-H), 7.35 (d, J = 7.5 Hz, 2H, Ar-H), 7.23 (d, J = 7.5 Hz, 1H, Ar-H), 2.60 (s, 3H, CH₃), 2.41 (d, J = 10.3 Hz, 9H, 3CH₃); ¹³C NMR (101 MHz, DMSO-d₆) δ_C (ppm) = 165.94, 164.58, 160.47, 142.25, 141.40, 139.24, 139.15, 132.34, 129.42, 128.31, 128.25, 127.46, 122.94, 120.31, 119.69, 25.54, 21.62, 21.51, 15.69, 15.00; Mass (m/z): 483 (M⁺, 38.50%), 79 (100%, base peak); Anal. Calcd. for C₂₇H₂₅N₅O₂S (483.59): C, 67.06; H, 5.21; N, 14.48. Found: C, 67.33; H, 5.38; N, 14.58%.

2.1.1.3. N-(4-((E)-1-(((Z)-5-Acetyl-3-(p-tolyl)-1,3,4-thiadiazol-2(3 H)-ylidene)hydrazono) ethyl) phenyl)-4-methylbenzamide 11c

Yellowish white powder (yield, 80%); m.p. = 205–207°C. FT-IR (v max, cm⁻¹): 3336, 3083, 3028, 2947, 2919, 1691, 1648, 1598; ¹H NMR (400 MHz, DMSO-d₆) δ_H (ppm) = 10.33 (s, 1H, NH), 7.90 (dd, J = 15.2, 6.2 Hz, 8H, Ar-H), 7.36 (dd, J = 13.6, 8.0 Hz, 4H, Ar-H), 2.58 (s, 3H, CH₃), 2.44–2.33 (m, 9H, 3CH₃); ¹³C NMR (101 MHz, DMSO-d6) δ_C (ppm) = 189.75, 165.48, 141.77, 131.87, 129.54, 128.95, 127.78, 127.00, 126.95, 122.11, 119.83, 25.04, 21.04, 20.63, 15.20; Mass (m/z): 483 (M⁺, 28.31%), 190 (100%, base peak); Anal. Calcd. for C₂₇H₂₅N₅O₂S (483.59): C, 67.06; H, 5.21; N, 14.48. Found: C, 57.25; H, 5.39; N, 14.55%.

2.1.2. General procedure for the synthesis of the final compounds 13a-c

The hydrazonoyl derivatives 12a-c underwent a reaction with an equal amount of compound 9 in 20 mL of absolute ethanol, employing 1 equivalent of triethylamine as a catalyst. All ingredients were heated under reflux for 6 h. Following the filtration and washing of the precipitates with hot ethanol, the final thiadiazole derivatives 13a-c were successfully obtained.

2.1.2.1. Ethyl (Z)-4-(3-chlorophenyl)-5-(((E)-1-(4-(4-methylbenzamido)phenyl) ethylidene) hydrazono)-4,5-dihydro-1,3,4-thiadiazole-2-carboxylate 13a

Yellow powder (yield, 72%); m.p. = 185–187°C. FT-IR (v max, cm⁻¹): 3338, 3309, 3061, 3029, 2981, 2920, 1716, 1671, 1657, 1600; ¹H NMR (400 MHz, DMSO-d₆) δ_H (ppm) = 10.33 (s, 1H, NH), 8.18 (t, J = 2.2 Hz, 1H, Ar-H), 7.97 (dd, J = 8.3, 2.1 Hz, 1H, Ar-H), 7.92–7.84 (m, 6H, Ar-H), 7.56 (t, J = 8.2 Hz, 1H, Ar-H), 7.42 (dd, J = 7.9, 2.0 Hz, 1H, Ar-H), 7.34 (d, J = 7.8 Hz, 2H, Ar-H), 4.37 (q, J = 7.1 Hz, 2H, CH₂), 2.40 (d, J = 6.1 Hz, 6H, 2CH₃), 1.33 (t, J = 7.1 Hz, 3H, CH₃); ¹³C NMR (101 MHz, DMSO-d₆) δ_C (ppm) = 166.01, 164.61, 161.22, 158.42, 142.13, 140.35, 139.94, 138.02, 133.78, 132.48, 131.34, 129.40, 129.17, 128.23, 127.16, 122.36, 122.28, 121.75, 120.56, 118.87, 63.41, 21.51, 16.05, 14.44; Mass (m/z): 534 (M⁺, 22.12%), 279 (100%, base peak); Anal. Calcd. for C₂₇H₂₄ClN₅O₃S (534.03): C, 60.73; H, 4.53; N, 13.11. Found: C, 60.88; H, 4.68; N, 13.36%.

2.1.2.2. Ethyl (Z)-5-(((E)-1-(4-(4-methylbenzamido)phenyl)ethylidene)hydrazono)-4-(p-tolyl)-4,5-dihydro-1,3,4-thiadiazole-2-carboxylate 13b

Yellowish white powder (yield, 70%); m.p. = 188–190°C. FT-IR (v max, cm⁻¹): 3338, 3062, 3028, 2980, 2918, 2862, 1745, 1717, 1671, 1648, 1600; ¹H NMR (400 MHz, DMSO-d₆) δ_H (ppm) = 10.34 (s, 1H, NH), 7.91–7.83 (m, 8H, Ar-H), 7.36 (t, J = 6.7 Hz, 4H, Ar-H), 4.39 (q, J = 7.1 Hz, 2H, CH₂), 2.39 (s, 6H, 2CH₃), 2.37 (s, 3H, CH₃), 1.33 (t, J = 7.1 Hz, 3H); ³C NMR (101 MHz, DMSO-d₆) δ_C (ppm) = 165.47, 163.40, 160.40, 157.92, 143.38, 141.76, 141.04, 139.94, 133.30, 131.98, 131.87, 130.80, 128.93, 127.79, 127.01, 126.49, 120.98, 119.77, 62.91, 21.04, 15.26, 13.95; Mass (m/z): 513 (M⁺, 22.93%), 492 (100%, base peak); Anal. Calcd. for C₂₈H₂₇N₅O₃S (513.62): C, 65.48; H, 5.30; N, 13.64. Found: C, 65.61; H, 5.42; N, 13.46%.

2.1.2.3. Ethyl (Z)-5-(((E)-1-(4-(4-methylbenzamido)phenyl)ethylidene)hydrazono)-4-(4-nitrophenyl)-4,5-dihydro-1,3,4-thiadiazole-2-carboxylate 13c

Yellow powder (yield, 70%); m.p. = 195–197°C. FT-IR (v max, cm⁻¹): 3338, 3103, 2977, 2918, 1744, 1657, 1599; ¹H NMR (400 MHz, DMSO-d₆) δ_H (ppm) = 10.34 (s, 1H, NH), 7.92–7.84 (m, 8H, Ar-H), 7.37–7.27 (m, 4H, Ar-H), 4.38 (q, J = 7.1 Hz, 2H, Ar-H), 2.39 (s, 6H, 2CH₃), 1.33 (t, J = 7.1 Hz, 3H, CH₃); ¹³C NMR (101 MHz, DMSO-d₆) δ_C (ppm) = 165.98, 161.84, 158.34, 144.99, 144.34, 142.27, 141.69, 132.32, 129.42, 128.26, 127.65, 125.43, 121.66, 120.28, 63.55, 21.51, 16.11, 14.44; Anal. Calcd. for C₂₇H₂₄N₆O₅S (544.59): C, 59.55; H, 4.44; N, 15.43. Found: C, 59.69; H, 4.75; N, 15.58%.

2.2. In vitro studies

2.2.1. 2.1. Cytotoxicity testing

The newly synthesized thiadiazole compounds, along with standard reference drugs, were tested for their effects on various cancer cell lines – MCF-7 (breast cancer), HepG-2 (liver cancer), and HCT-116 (colon cancer) – as well as a normal human fibroblast cell line (WI-38). The MTT assay was used to evaluate cell survival [51,52]. Full experimental details can be found in the Supplementary Materials.

2.2.2. VEGFR-2 inhibition

Compound 11a and sorafenib (used as a control inhibitor) were assessed for their ability to block VEGFR-2 activity. This experiment was done using enzyme assay kits at concentrations from 0.3 to 1000 nM [53]. The full procedure is described in the Supplementary Materials.

2.2.3. Flow cytometry: apoptosis and cell cycle analysis

Flow cytometry was used to analyze how compound 11a affects programmed cell death (apoptosis) and changes in the cell cycle of MCF-7 cells. Detailed steps are available in the Supplementary Materials [54].

2.2.4. Apoptosis-related protein expression

The effects of compound 11a on apoptosis-related proteins – caspase-3, Bcl-2, and Bax – were investigated in MCF-7 cells [55]. Methods and conditions are described in detail in the Supplementary Materials.

2.2.5. Statistical analysis section

GraphPad Prism 6 software (GraphPad, San Diego, CA, USA) was used to analyze all results. The data were represented as mean ± standard deviation (Mean ± SD) from at least three independent experiments (n = 3). One-way analysis of variance (One-way ANOVA) followed by Tukey post hoc multiple comparison tests were elected to analyze the significant difference between all groups. p < 0.05 was considered statistically significant.

2.3. Computational studies

2.3.1. Molecular docking

Computer-based docking was conducted to explore how the thiadiazole compounds and reference drugs fit into the VEGFR-2 active site [56]. More details are provided in the Supplementary Materials.

2.3.2. Molecular dynamics simulations

To assess how stable the complex between VEGFR-2 and compound 11a is over time, molecular dynamics simulations were carried out using CHARMM-GUI and GROMACS 2021 [57,58]. Binding energies were calculated using the MM-GBSA method via the gmx_MMPBSA tool [59,60]. Principal Component Analysis (PCA) was also used to track movement in key amino acids [61]. More details are provided in the Supplementary Materials.

2.3.3. ADMET and toxicity prediction

Using Discovery Studio 4.0, computer models predicted how the thiadiazole compounds behave in the body and their potential toxicity, providing insight into their drug-likeness and safety [62].

2.3.4. Electronic properties via DFT

Compound 11a’s structure and electronic behavior were studied using Density Functional Theory (DFT) with Gaussian 09 software [63]. The calculations used the B3LYP method and 6–31+G(d,p) basis set to analyze molecular stability and reactivity. More details are provided in the Supplementary Materials.

3. Results and discussions

3.1. Chemistry

This research focuses on the synthesis of newly developed thiadiazole compounds having the essential pharmacophoric features of VEGFR-2 inhibitors. The compounds were synthesized following the pathways illustrated in Figure 2 (Scheme 1). Compound 7 was synthesized by adding 4-methyl benzoyl chloride 6 to 4-aminoacetophenone 5 while stirring in an ice bath. The compound 7 was then refluxed with methyl hydrazinecarbodithioate 8 in absolute ethanol, resulting in methyl (E)-2-(1-(4-(4-methylbenzamido)phenyl)ethylidene)hydrazine-1-carbodithioate 9. Ultimately, compound 9 reacted with the hydrazonyl derivatives 10a-c and 12a-c in ethanol under reflux conditions, using triethylamine as a catalyst, leading to the formation of the desired thiadiazole derivatives 11a-c and 13a-c, respectively.

Figure 2.

Scheme 1. Synthesis of the final compounds 11a-c and 13a-c.

The structural elucidation of the synthesized thiadiazole derivatives (11a-c and 13a-c) was thoroughly validated through the use of various spectroscopic techniques, such as FT-IR, ¹H and ¹³C NMR, and mass spectrometry. The FT-IR spectra of all synthesized derivatives displayed characteristic bands consistent with the expected functional groups. Notably, a sharp absorption band around 3336–3340 cm⁻¹ was observed, corresponding to the NH stretching vibration, which is indicative of the presence of the amide and hydrazone groups. Additionally, strong carbonyl (C = O) stretching bands were recorded in the range of 1657–1717 cm⁻¹. These carbonyl absorptions confirmed the amide, acetyl, or ester functionalities within the molecular framework. Aromatic C-H stretching vibrations appeared in the region of 3060–3083 cm⁻¹, while alkyl C-H stretching bands were observed around 2918–2966 cm⁻¹, reflecting the methyl and ethyl substituents present in the molecules. The ¹H NMR spectra offered precise information on the proton environments of the produced molecules. All derivatives showed a singlet at δ_H = 10.3 ppm, which corresponds to the NH proton of the amide or hydrazone group. Aromatic protons were detected in multiplets at δ_H = 7.9–7.3 ppm, confirming the compounds’ aromatic backbone. The methyl groups connected to aromatic or acetyl substituents exhibited as singlets or doublets at δ_H = 2.2–2.6 ppm. In derivatives with ethoxy groups, signals for CH₂ and CH₃ protons were recognized as a quartet (δ_H ~4.4 ppm) and a triplet (δ_H ~1.3 ppm), respectively. The ¹³C NMR spectra corroborated the molecular structure by displaying distinctive carbon signals. Carbonyl carbons were seen at δ_C ~165–190 ppm, contingent upon their association with an amide, acetyl, or ester functional group. Aromatic carbons were detected within the δ_C = 120–145 ppm range, whereas methyl carbons were noted at around δ_C 14–25 ppm. The mass spectrometric analysis provided additional confirmation of the molecular weights and fragmentation patterns of the synthesized compounds. The CHN elemental analysis established the composition of the produced compounds by determining the percentages of carbon, hydrogen, and nitrogen in the molecules. The experimental results were quite similar to the theoretical values derived for the chemical formulae of the target derivatives. This agreement gave further assurance that the chemicals were pure and successfully synthesized.

3.2. In vitro evaluations

3.2.1. Cytotoxicity

The in vitro cytotoxicity of the synthesized thiadiazole derivatives (compounds 11a–13c) was assessed against a panel of human cancer cell lines – MCF-7 (breast adenocarcinoma), HepG-2 (hepatocellular carcinoma), and HCT-116 (colon carcinoma) – along with WI-38, a normal human lung fibroblast cell line, to evaluate potential anticancer activity and selectivity (Table 1). The clinically approved VEGFR-2 inhibitor, sorafenib, was included as a reference compound for comparative purposes. The IC₅₀ values revealed that compound 11a exhibited the most promising anticancer potential among the tested series. It showed cytotoxicity in the low micromolar range across the MCF-7 (9.49 µM), HepG-2 (12.89 µM), and HCT-116 (28.97 µM) cell lines, indicating broad-spectrum activity. Although sorafenib demonstrated superior potency, particularly in HCT-116 cells (6.57 µM), compound 11a’s reduced cytotoxicity in normal cells (WI-38, 41.36 µM) suggests a favorable selectivity index and potential for therapeutic window optimization.

Table 1.

In vitro cytotoxicity, selectivity profiles, and p-values vs. Sorafenib in cancerous and normal cell lines.

Comp.	In vitro cytotoxicity IC50 (µM)*, selectivity indexes, and p-values
	WI-38	MCF-7	SI	p-value	HepG-2	SI	p-value	HCT-116	SI	p-value
11a	41.36 ± 2.5	9.49 ± 0.7	4.36	0.041	12.89 ± 0.9	3.21	0.035	28.97 ± 1.7	1.43	0.029
11b	65.24 ± 3.7	39.59 ± 2.4	1.65	0.050	49.03 ± 2.9	1.33	0.042	52.10 ± 3.2	1.25	0.037
11c	32.82 ± 2.1	63.71 ± 3.6	0.52	0.018	76.38 ± 3.8	0.43	0.021	81.25 ± 4.1	0.40	0.016
13a	22.68 ± 1.8	54.35 ± 3.2	0.42	0.009	61.58 ± 3.4	0.37	0.011	87.71 ± 4.3	0.26	0.007
13b	44.52 ± 2.6	76.22 ± 3.9	0.58	0.002	83.13 ± 4.2	0.54	0.003	91.96 ± 5.1	0.48	0.002
13c	16.23 ± 1.4	46.88 ± 2.8	0.35	0.047	34.92 ± 2.2	0.46	0.043	66.67 ± 3.5	0.24	0.039
Sorafenib	11.75 ± 0.8	8.16 ± 0.3	1.44	–	10.28 ± 0.6	1.14	–	6.57 ± 0.3	1.79	–

*Values are given as mean ± SEM of three independent experiments (n = 3). Statistical significance was evaluated using one-way ANOVA followed by Tukey post hoc post hoc test (each compound vs. sorafenib). p < 0.05: statistically significant; p < 0.01: moderately significant; p < 0.001: highly significant.

To quantitatively assess selectivity, the selectivity index (SI) was calculated as the ratio of IC₅₀ in WI-38 cells to IC₅₀ in cancer cell lines. For compound 11a, the SI values were 4.36 (MCF-7), 3.21 (HepG-2), and 1.43 (HCT-116), reflecting its most favorable selectivity for breast and liver cancer lines. In contrast, sorafenib, despite being more potent overall, demonstrated SI values of 1.44 (MCF-7), 1.14 (HepG-2), and 0.56 (HCT-116), indicating lower cancer-to-normal cell selectivity compared to compound 11a. This underscores 11a’s potential for a better therapeutic window, particularly for breast and liver cancers, and suggests that further optimization may yield even more selective agents.

Other compounds in the series exhibited markedly lower selectivity. For example, compound 11b, with modest activity against cancer cells (IC₅₀ > 39 µM across all lines), showed SI values barely above 1.0, indicating no preferential cytotoxicity and thus limited therapeutic relevance. Compound 13c, which exhibited activity in HepG-2 (34.92 µM) and MCF-7 (46.88 µM), yielded SI values of 0.47 and 0.35, respectively, due to its relatively higher toxicity toward WI-38 (16.23 µM), raising concerns about off-target effects.

In conclusion, compound 11a emerges as the most promising thiadiazole derivative in this series, combining moderate potency with high selectivity, especially in MCF-7 and HepG-2 cell lines. While it does not surpass the potency of sorafenib, its selectivity indices suggest a superior safety profile. These findings provide a compelling rationale for further optimization of the thiadiazole scaffold and subsequent in vivo validation of compound 11a as a potential VEGFR-2-targeted anticancer agent.

3.2.2. Structure–activity relationship (SAR)

The cytotoxic activity of compounds 11a–11c and 13a–13c was evaluated against MCF-7, HepG-2, and HCT-116 cancer cell lines, using Sorafenib as a reference drug. Lower IC₅₀ values indicate higher cytotoxic potency.

Effect of electronic nature and substitution position on the phenyl ring at the 3-position of the thiadiazole ring: The position and electronic character of substituents on the phenyl ring at the 3-position of the thiadiazole ring played a significant role in modulating anticancer activity across the cell lines tested.

• Comparison of 11b and 11c (meta- versus para-substituted electron-donating group): Compound 11b (3-CH₃, meta-substituted EDG) showed better cytotoxicity than 11c (4-CH₃, para-substituted EDG) across all cell lines. Specifically, 11b had IC₅₀ values of 39.59 µM (MCF-7), 49.03 µM (HepG-2), and 52.10 µM (HCT-116), while 11c showed significantly weaker activity with IC₅₀ values of 63.71 µM (MCF-7), 76.38 µM (HepG-2), and 81.25 µM (HCT-116). This suggests that meta-substitution with EDGs enhances cytotoxic activity, likely by offering a more favorable orientation or interaction with the target site.

• Comparison of 13b and 13c (para-substituted EDG versus EWG): Both 13b and 13c feature para-substituted phenyl rings, differing only in electronic properties—13b carries a 4-CH₃ (EDG), while 13c has a 4-NO₂ (EWG). 13c demonstrated superior cytotoxicity with IC₅₀ values of 46.88 µM (MCF-7), 34.92 µM (HepG-2), and 66.67 µM (HCT-116), compared to 13b, which had 76.22 µM (MCF-7), 83.13 µM (HepG-2), and 91.96 µM (HCT-116). These results indicate that EWGs at the para position improve biological activity more than EDGs in this scaffold.

• Comparison of 11c and 13b (acetyl versus ethyl formate, same para-CH₃): Despite both compounds sharing a 4-CH₃ para-substituent, 11c (acetyl at the 2-position) consistently outperformed 13b (ethyl formate at the 2-position). 11c had IC₅₀ values of 63.71 µM (MCF-7), 76.38 µM (HepG-2), and 81.25 µM (HCT-116), while 13b had 76.22 µM (MCF-7), 83.13 µM (HepG-2), and 91.96 µM (HCT-116). This indicates that the acetyl group provides a more favorable pharmacophore than ethyl formate in the presence of an EDG at the 3-position.

3.2.3. Inhibition of VEGFR-2

To evaluate the biological effectiveness of the lead compound 11a, its inhibitory activity was compared with sorafenib, a well-established and clinically approved VEGFR-2 inhibitor. Remarkably, compound 11a demonstrated an IC₅₀ value of 0.055 ± 0.004 µM, which is nearly equivalent to that of Sorafenib (0.043 ± 0.005 µM), indicating a comparable level of potency. This comparable IC₅₀ value emphasizes compound 11a’s ability to exert strong VEGFR-2 inhibition at low micromolar concentrations. Furthermore, it effectively induces cytotoxic effects, suggesting its potential for therapeutic efficacy at clinically meaningful dosage levels. Compound 11a’s structural novelty combined with potent activity positions it as a compelling candidate for further preclinical investigation and optimization in the pursuit of safer and more effective anti-cancer agents targeting VEGFR-2.

3.2.4. Comparative benchmarking of compound 11a against clinical VEGFR-2 inhibitors

To contextualize the efficacy of compound 11a, we benchmarked its VEGFR-2 inhibition profile against established clinical VEGFR-2 inhibitors. Compound 11a demonstrated an in vitro VEGFR-2 enzymatic IC₅₀ of 0.055 ± 0.004 µM, which is comparable to sorafenib (IC₅₀ = 0.043 ± 0.005 µM, measured under the same conditions). This positions 11a within the low nanomolar potency range, consistent with other clinically used VEGFR-2 inhibitors such as lenvatinib (IC₅₀ = 4–6 nM) and tivozanib (IC₅₀ = 0.2–1.6 nM) as reported in the literature [64,65].

While 11a may be less potent than tivozanib or lenvatinib on a per-molecule basis, its potency is still within a therapeutically relevant window, particularly when considering its favorable selectivity indices (SI > 3 in MCF-7 and HepG-2) and reduced cytotoxicity toward normal cells. Furthermore, molecular docking and 200 ns MD simulations confirmed a stable interaction profile of 11a within the VEGFR-2 active site, engaging key residues such as Glu883 and Asp1044, similar to the interaction hotspots exploited by known type II kinase inhibitors.

Taken together, these data support compound 11a as a valid lead candidate, with competitive VEGFR-2 inhibition relative to FDA-approved drugs, and potential for further optimization to improve its potency and drug-like properties.

3.2.5. Analysis of MCF-7 cells’ cycle

To further elucidate the mechanism of anti-proliferative activity, the effect of compound 11a on cell cycle progression was assessed in MCF-7 cells using flow cytometry, and the results were compared to those of untreated control cells (Table 2 and Figure 3(A)). The distribution of cell populations across the G0/G1, S, and G2/M phases revealed a significant alteration in the cell cycle profile upon treatment with compound 11a. In untreated control cells, the majority of cells were in the G0/G1 phase (38.02%), followed by S phase (35.52%) and G2/M phase (26.46%), indicating a normal proliferative state.

Table 2.

Impact of compound 11a on cell cycle progression in MCF-7 breast cancer cells.

Sample	G0/G1	S	G2/M
Comp. 11a	13.34*	43.86	42.8*
Cont. MCF7	38.02	35.52	26.46

*significant p value; significant p value < 0.05 & by using One-way ANOVA followed by Tukey’s post hoc multiple comparison tests.

Figure 3.

Flow cytometry analysis: A) Effect of compound 11a on cell cycle distribution in MCF-7 breast cancer cells. B) Apoptosis and necrosis rates of MCF-7 cells treated with compound 11a.

In contrast, treatment with compound 11a markedly decreased the G0/G1 population to 13.34%, while inducing a substantial accumulation of cells in the S phase (43.86%) and G2/M phase (42.8%). This dual accumulation strongly suggests that compound 11a disrupts normal cell cycle progression by inhibiting both DNA synthesis and mitotic entry, leading to a cell cycle arrest at S and G2/M phases. Such a shift typically correlates with DNA damage response activation or inhibition of cell cycle checkpoint regulators, pointing toward the compound’s potential to interfere with cancer cell proliferation via replication and division blockade. The observed S and G2/M phase arrest aligns well with the compound’s cytotoxic profile and further supports its mechanism as a potential VEGFR-2-targeted anti-proliferative agent.

3.2.6. Analysis of MCF-7 cells’ apoptosis and necrosis rates

To further confirm the mode of cell death induced by compound 11a, an apoptosis assay was conducted using annexin V/PI staining followed by flow cytometry analysis in MCF-7 cells (Table 3 and Figure 3(B)). The results revealed a dramatic shift in cell viability and death profile compared to untreated control cells. In the control group, a high percentage of cells (93.79%) remained viable, with only 2.13% in early apoptosis, 3.67% in late apoptosis, and 0.41% undergoing necrosis. This profile is indicative of a healthy, proliferative cell population.

Table 3.

Induction of apoptosis and necrosis in MCF-7 cells by compound 11a.

Sample	Viable cells	Early apoptosis	Late apoptosis	Necrosis
Comp. 11a	27.46*	37.75*	32.68*	2.12*
Cont. MCF7	93.79	2.13	3.67	0.41

*significant p value; significant p value < 0.05 & by using One-way ANOVA followed by Tukey’s post hoc multiple comparison tests.

Upon treatment with compound 11a, the percentage of viable cells sharply decreased to 27.46%, while cells in early apoptosis rose to 37.75% and those in late apoptosis to 32.68%. This clearly demonstrates that compound 11a induces a significant apoptotic response, accounting for over 70% of the total cell population. The necrotic fraction remained minimal (2.12%), suggesting that compound 11a predominantly induces programmed cell death, rather than nonspecific membrane lysis or necrosis. Such a profile is highly desirable in anticancer therapy, as apoptosis avoids the inflammatory responses typically associated with necrosis.

These findings, together with the observed cell cycle arrest in S and G2/M phases, underscore the compound’s ability to halt cancer cell proliferation and trigger apoptosis in a controlled, non-necrotic manner. Combined with its favorable selectivity index and in vitro potency, compound 11a appears to exert multifaceted anticancer activity, supporting its potential for further preclinical development.

3.2.7. Key apoptotic proteins assay

To elucidate the molecular mechanisms underlying the pro-apoptotic activity of compound 11a, the expression levels of key apoptotic proteins (caspase-3, Bcl-2, and Bax) were quantified in MCF-7 cells. Compared to untreated control cells, treatment with compound 11a led to a profound upregulation of caspase-3, a critical executioner caspase in the apoptotic cascade. Specifically, caspase-3 concentration increased from 55.77 ng/mL in control cells to 371.49 ng/mL in treated cells, representing a 6.66-fold elevation. This suggests robust activation of the execution phase of apoptosis following mitochondrial damage (Table 4).

Table 4.

Expression levels of apoptotic markers caspase-3, Bcl-2, and Bax after treatment with compound 11a.

Sample	Caspase-3 conc.	FLD	Bcl-2 conc.	FLD	Bax conc.	FLD
11a	371.49*	6.66	2.04*	2.67	363.29*	6.89
Cont. MCF7	55.77	1	5.45	1	52.75	1

*significant p value; significant p value < 0.05 & by using One-way ANOVA followed by Tukey’s post hoc multiple comparison tests.

In parallel, the anti-apoptotic protein Bcl-2 was significantly downregulated from 5.45 ng/mL to 2.04 ng/mL, corresponding to a 2.67-fold decrease, while the pro-apoptotic protein Bax was markedly upregulated from 52.75 ng/mL to 363.29 ng/mL, showing a 6.89-fold increase. The resulting high Bax/Bcl-2 ratio is a well-established trigger for mitochondrial outer membrane permeabilization, leading to cytochrome c release and subsequent caspase activation.

Collectively, these alterations confirm that compound 11a induces apoptosis through activation of the intrinsic mitochondrial pathway, as evidenced by elevated Bax and caspase-3 levels alongside Bcl-2 suppression. This molecular signature complements the earlier flow cytometry findings, which showed increased early and late apoptotic populations, and aligns with the observed cell cycle arrest at S and G2/M phases. The mechanistic consistency across phenotypic and molecular data strengthens the candidacy of compound 11a as a targeted anticancer agent acting via mitochondrial-mediated apoptosis.

3.3. In silico evaluations

3.3.1. Molecular docking

The docking of compound 11a, identified as the most potent molecule in this study, within the VEGFR-2 active site exhibited an optimal binding orientation that improves its interaction potential. The molecule exhibited significant stability through the formation of two hydrogen bonds with the essential amino acids, Glu883 and Asp1044. These interactions are crucial for ensuring the stable positioning of compound 11a within the binding pocket, thereby enhancing its binding affinity and efficacy. The hydrogen bond with Glu883 is crucial, as this residue plays a key role in maintaining the structural integrity of the VEGFR-2 active site. The interaction with Asp1044 enhances the compound’s binding by contributing to a highly specific and complementary binding environment for the ligand. The interactions strengthen the stability of compound 11a within the VEGFR-2 active site and highlight its potential as an effective inhibitor of the receptor. Compound 11a‘s ability to form multiple critical interactions with these amino acids suggests its potential to inhibit receptor function, thereby interfering with the angiogenesis and tumor proliferation processes mediated by VEGFR-2. This profile identifies compound 11a as a promising candidate for further development in cancer therapies, particularly in targeting the VEGF signaling pathway, essential for tumor vascularization and progression (Figure 4).

Figure 4.

Compound 11a within the VEGFR-2 active site, shown in 3D (top) and 2D (bottom).

3.3.2. MD simulations

Molecular dynamics simulation is a pivotal tool in rational drug design as it enables the real-time assessment of ligand – receptor interactions under physiologically relevant conditions. By modeling atomic motions over extended timescales, MD allows us to evaluate binding stability beyond static docking poses, revealing whether a ligand remains stably accommodated within its target site. It also provides detailed insight into protein and ligand flexibility, conformational rearrangements, and solvation effects, which are critical for understanding binding mechanisms and predicting functional outcomes. Quantitative analyses such as root mean square deviation (RMSD) for global stability, root mean square fluctuation (RMSF) for residue-level flexibility, and solvent-accessible surface area (SASA) for solvation changes can be complemented by free energy estimation methods like MM-PBSA to rank ligand affinities. Together, these tools offer a comprehensive dynamic and energetic framework to guide the optimization of lead compounds.

The temporal evolution of protein backbone RMSD was strictly examined across the simulation trajectories for both the VEGFR-2-11a holo complex and the apo VEGFR-2 system (Figure 5(A)). Initial equilibration phases revealed a rapid ascent in RMSD values within the first 20 ns, indicative of structural relaxation and adaptation from the energy-minimized starting conformations. Subsequent trajectory analysis highlighted distinct divergence in stability profiles between the holo and apo systems. The VEGFR-2-11a holo system exhibited a relatively stable RMSD profile, fluctuating between approximately 2.5 and 3.5 Å throughout the 200 ns simulation. This behavior underscores a high degree of intrinsic structural stability in ligand binding, suggesting that VEGFR-2 maintains a robust conformational integrity while accommodating 11a within its binding domain.

Figure 5.

MD simulation analyses for compound 11a-VEGFR-2 complex. (A) RMSD of VEGFR-2 Cα atoms over time, comparing holo and apo simulations. Holo VEGFR-2 (red) maintains lower RMSD fluctuations relative to the apo form (green), indicating enhanced structural stability upon ligand binding. (B) RMSF per residue for VEGFR-2 Cα atoms. Both states exhibit similar flexibility profiles, with notable peaks corresponding to flexible loop regions. (C) Rg of VEGFR-2 as a function of time. Holo VEGFR-2 demonstrates a more compact and stable global structure than the apo form. (D) SASA evolution over time. Minor differences in SASA values suggest a slightly more compact surface exposure profile in the holo complex.

Conversely, the apo VEGFR-2 system demonstrated a progressive increment in RMSD values, particularly beyond the 100 ns mark, reaching maxima in the 4.0 to 6.0 Å range. This substantial deviation reflects the intrinsic conformational lability of the unbound VEGFR-2 receptor. Notably, the shaded regions enveloping each RMSD trajectory represent the standard deviation calculated across three independent replicates, providing a rigorous statistical estimation of structural variability inherent to each system. These findings suggest that the absence of ligand-stabilizing interactions provokes broader conformational exploration and dynamic plasticity, likely facilitating receptor flexibility associated with basal functional states. Such differential RMSD behavior between holo and apo states not only substantiates the stabilizing effect of 11a binding on VEGFR-2 structural dynamics but also reinforces the broader theoretical framework whereby ligand engagement acts as a molecular constraint, tempering excessive conformational entropy, and promoting functional specificity in receptor behavior.

The per-residue Root Mean Square Fluctuation (RMSF) profiles further elucidated the local flexibility patterns underlying the global RMSD observation (Figure 5(B)). Both the holo and apo systems displayed elevated RMSF values at the N- and C-terminal regions, consistent with the intrinsic disorder typically observed in terminal segments of proteins. Beyond the termini, notable flexibility peaks were observed between residues 1045 and 1065, corresponding to loop regions proximal to the 11a binding site. Notably, the apo VEGFR-2 system consistently demonstrated higher RMSF magnitudes within these loop regions than the holo counterpart, indicating greater conformational freedom without ligand anchoring interactions. The VEGFR-2-11a holo system exhibited a discernible reduction in local fluctuations across key binding interface residues, underscoring the stabilizing influence exerted by ligand engagement. The shaded areas surrounding the RMSF traces represent the standard deviations across three independent replicates, emphasizing the robustness of these differential flexibility patterns. These RMSF findings harmonize coherently with the RMSD observations, reinforcing the interpretation that 11a binding constrains VEGFR-2 dynamic excursions, thereby imparting localized rigidity while preserving global structural stability. This nuanced interplay between regional flexibility and overall conformational control is critical for functional modulation of VEGFR-2 signaling upon ligand binding.

The analysis of the radius of gyration (Rg) provided complementary insights into the global compactness and tertiary structural integrity of the VEGFR-2 protein across holo and apo states (Figure 5(C)). Throughout the 200 ns trajectory, the VEGFR-2-11a holo system consistently exhibited slightly higher Rg values than the apo VEGFR-2, indicating a modest yet persistent expansion of the molecular framework upon ligand binding. Specifically, the holo system maintained an average Rg around ~ 20.6 Å, whereas the apo counterpart demonstrated a marginally lower average Rg near ~ 20.4 Å. The shaded regions, representing standard deviations across three independent replicates, further corroborate these trends by highlighting the low variability in Rg values, particularly within the holo system.

These findings suggest that the binding of 11a induces subtle structural dilation within VEGFR-2, likely reflecting localized loosening of the tertiary structure required to accommodate the ligand. This slight expansion is consistent with the increased flexibility observed in specific regions by RMSF analysis, yet does not compromise the overall conformational stability as confirmed by the relatively stable RMSD trajectories. Critically, this minor increase in Rg may hold functional significance by modulating receptor activation dynamics, potentially facilitating allosteric signal transduction mechanisms that are known to depend on fine-tuned structural adaptability in receptor tyrosine kinases such as VEGFR-2.

The Solvent Accessible Surface Area (SASA) profiles provided critical complementary information regarding the exposure of the VEGFR-2 surface to the solvent environment across the holo and apo systems (Figure 5(D)). Throughout the simulation, the VEGFR-2-11a holo complex consistently exhibited higher SASA values than the apo VEGFR-2, indicative of enhanced solvent exposure upon ligand binding. Quantitatively, the holo structure maintained an average SASA of around ~17,500 Ų, while the apo structure demonstrated a slightly reduced SASA, averaging ~17,200 Ų. The standard deviations represented by the shaded areas confirmed that these differences were systematically maintained across three independent replicates.

This observed elevation in SASA for the holo system aligns coherently with the slight increase in radius of gyration, reinforcing the interpretation that 11a binding induces a subtle yet discernible expansion and partial loosening of the VEGFR-2 molecular structure. From a functional perspective, such an increase in solvent exposure could modulate the accessibility of regulatory regions or allosteric sites, thereby influencing receptor activation states or downstream signaling efficacy. Although modest, the consistent increase in solvent exposure underscores the profound impact that small-molecule ligands like 11a can exert on the conformational landscape and biophysical properties of membrane-associated receptor kinases such as VEGFR-2.

The dynamic profiles of intramolecular hydrogen bond formation were systematically evaluated across the VEGFR-2-11a holo complex and apo VEGFR-2 systems to assess internal stabilization and secondary structure integrity (Figure 5(E)). Throughout the 200 ns simulations, both systems exhibited relatively stable hydrogen bond counts, fluctuating within approximately 60 to 75 hydrogen bonds. The VEGFR-2-11a holo and the apo demonstrated highly comparable hydrogen bonding patterns, with only marginal differences observed. The apo system tended to exhibit slightly higher instantaneous hydrogen bond numbers at isolated time points; however, the overall average values remained statistically indistinguishable within the bounds of the shaded standard deviation regions derived from three independent replicates.

Figure 5.

(E) Number of intramolecular hydrogen bonds formed during the simulation. No significant differences were observed between the holo and apo states, supporting a comparable overall folding stability. (F) 11a RMSD relative to the initial structure during the holo simulations. 11a shows moderate positional fluctuations but remains generally stable within the binding pocket. (G) Distance between the center of mass of VEGFR-2 and the center of mass of 11a. The distance stabilizes throughout the simulation, indicating persistent binding of the ligand to VEGFR-2 with minimal dissociation events.

This consistency in hydrogen bond counts across holo and apo states suggests that 11a binding does not significantly disrupt the intrinsic hydrogen bonding network that underpins VEGFR-2’s secondary and tertiary structure. Instead, ligand engagement modulates local flexibility and solvent exposure as evidenced by RMSF and SASA analyses. These findings further corroborate the interpretation that 11a stabilizes VEGFR-2 through subtle modulation of dynamic properties rather than gross structural perturbation, a mechanistic nuance of particular importance for ligand design strategies aiming to fine-tune receptor activity while preserving native fold stability.

The dynamic stability of the 11a ligand within the VEGFR-2 binding pocket was rigorously assessed through analysis of both ligand RMSD relative to the initial bound pose and center of mass (COM) distances between the ligand and the receptor (Figure 5(F,G)). Ligand RMSD profiles revealed moderate fluctuation during the first 50 ns, with RMSD values stabilizing between approximately 4.5 and 5.5 Å. While intermittent excursions beyond 6.0 Å were observed at specific time points, the ligand consistently returned to a stable RMSD range, indicating no dissociation events and suggesting dynamic yet persistent accommodation within the binding site. Notably, the shaded areas representing standard deviation across replicates demonstrated relatively modest variability, reinforcing the robustness of these observations. Complementary COM distance analyses further substantiated these findings. Throughout the trajectory, the COM distance between 11a and VEGFR-2 remained stable around ~10 to 11 Å, with only minor oscillations and no evidence of progressive drift indicative of ligand unbinding or substantial repositioning.

Collectively, these ligand-centric parameters corroborate the holo protein observations wherein 11a binding induced minor expansions in global compactness (as evidenced by increased Rg and SASA) while maintaining overall structural coherence. The persistent binding of 11a, characterized by stable RMSD and COM profiles, suggests that ligand engagement enforces localized conformational remodeling without compromising the anchoring interactions necessary for binding stability. These data support a model wherein 11a exerts a fine-tuning effect on VEGFR-2 dynamics, stabilizing key functional regions while permitting adaptive plasticity within the broader conformational ensemble.

Comprehensive principal component analysis (PCA), conformational landscape mapping, binding pocket plasticity, and detailed interaction fingerprinting of the VEGFR-2–compound 11a complex – supported by multi-replicate clustering and MMGBSA free energy decomposition – are thoroughly presented in the Supplementary Materials (S3.4.1). Our MD simulations reveal that 11a binding not only stabilizes the VEGFR-2 structure, reducing global and local flexibility compared to the apo state, but also constricts the receptor’s conformational ensemble toward functionally poised states. Interaction fingerprinting and clustering analyses identify a robust pharmacophoric anchor, with reproducible hydrogen bonding and hydrophobic contacts underpinning persistent ligand engagement across trajectories (S3.5.1).

Critically, MMGBSA free energy calculations provide atomic-level insight into the energetic landscape of ligand binding (S3.3.1 and Fig. S3.3.1.1). Per-residue decomposition pinpoints a set of key residues – such as Met911, Glu915, Ile913, Phe1045, and Glu883—as dominant contributors to binding stability, in agreement with interaction maps from the trajectory (Fig. S3.3.1.2). The overall binding free energy of −35.07 kcal/mol underscores the high affinity and specificity of 11a for VEGFR-2. Component analysis further demonstrates that ligand binding is driven by strong van der Waals and electrostatic interactions, partially counteracted by the desolvation penalty, reflecting the classical hallmarks of potent kinase inhibitor binding. Collectively, these findings provide compelling atomic-level evidence for the stability, adaptability, and selectivity of 11a–VEGFR-2 interactions. Readers are encouraged to explore the Supplementary Materials for in-depth figures, per-residue energy maps, trajectory analyses, and methodological details that substantiate these mechanistic insights and highlight their significance for rational inhibitor design.

3.3.3. Density functional theory (DFT)

DFT simulations were used to assess the prospective candidate’s electronic characteristics, including electron density, energy gap, and molecular orbital distribution (HOMO and LUMO). In order to shed light on the candidate’s stability, chemical reactivity, and possible manner of binding with the protein target, all relevant HOMO/LUMO characteristics are computed. The B3LYB approach was applied with 6-31 G+(d,p) using Gaussian 09‘s default parameter values for the DFT studies. We calculated the quantum global reactivity descriptors and mainly analyzed the energy gap, the atomic Mulliken charge distribution, the surface electrostatic potential, and the total density of states (TDOS). GaussView 5.0 and GaussSum 3.0 were used to visualize and evaluate the resulting file after compound 11a was optimized and relaxed. Fig. S3.6.1A is the labeled optimized structure, and Table S3.6.2 contains the dipole moment (Dm) and ground stated energy that were computed. As seen in Fig. S3.6.1A, the two terminal benzene rings on the planar portion are observed to be slanted. Because of the Dm value of 7.253 Debye, the molecule has a rather sizable asymmetric charge distribution, which promotes intermolecular interactions like hydrogen bonding.

The HOMO and LUMO functions distribute almost exclusively over the planar portion of the compound, but the lobes of HOMO function are large as shown in Fig. S3.6.1B. With smaller density lobes in the case of LUMO functions plus the weak overlap between the pi orbitals of the tilted terminal benzene rings and the flat portion of the complex, the molecule’s ability to take electrons is reduced. Thus, the donating capabilities to the target protein would be greater than the accepting properties. Since the intensity in the HOMO case is higher than those in the LUMO scenario, the higher TDOS intensity supports this observation (Fig. S3.6.1C). The atomic Mulliken charge analysis highlights the asymmetric charge distribution shown by the Dm value, as shown in Fig. S3.6.1D. The color scale indicates that the green C14 is the most positive atomic site and the red C26 is the most negative. While electrophilic attacks take place at negative active sites, nucleophilic attacks take place at the positive sites.

A comparatively low gap energy of 3.271 eV was observed as the gap value between HOMO and LUMO. Higher chemical reactivity and less kinetic stability are associated with a smaller energy gap, or E_gap. The HOMO/LUMO energy was used to determine the descriptors in Table S3.6.2, and the results indicated that the molecule is soft and may have inhibitory reactivity. The surface electrostatic potential controls the candidate’s chemical reactivity toward 11a in terms of both electrophilic and nucleophilic assaults. The distribution of nucleophilic and electrophilic centers can be inferred from the electrostatic potential. The negative potential areas that are advantageous for the target’s electrophilic sites are indicated in red on the electrostatic potential map in Fig. S3.6.1E. Over oxygen atoms, two red batches are concentrated. Negatively charged locations at the target benefit from the blue batches concentrating on hydrogen atoms. Because of their neutral charge, the green areas are conducive to hydrophobic interactions

3.3.4. In silico ADMET

To assess the pharmacokinetic and drug-likeness properties of the synthesized thiadiazole derivatives, an in silico ADMET analysis was conducted (Table 5), focusing on key parameters including blood–brain barrier (BBB) permeability, solubility, intestinal absorption, CYP2D6 inhibition, and plasma protein binding (PPB). All compounds, including sorafenib, demonstrated a “very low” BBB penetration potential, which is advantageous for minimizing central nervous system (CNS)-related side effects in anticancer agents that do not require CNS targeting.

Table 5.

Computer-based ADMET analysis of thiadiazole compounds and sorafenib.

Comp.	BBB level	Solubility level	Absorption level	CYP2D6 prediction	PPB prediction	Caco-2 Permeability	ALog P
11a	Very low	Very low	Poor	Non inhibitor	More than 90%	0.5196	5.895
11b			Moderate			0.5344	5.717
11c						0.5344	5.717
13a			Poor			0.5648	6.369
13b						0.5421	6.191
13c						0.5617	5.599
Sorafenib			Good			0.5138	4.175

The solubility predictions uniformly categorized all compounds, including the reference drug, as “very low,” suggesting potential challenges in aqueous formulation and oral delivery. This limitation may be addressed through prodrug strategies, nanoformulations, or co-solvent systems in future development. Regarding intestinal absorption, compounds 11b and 11c showed moderate predicted absorption, while others, including the lead compound 11a and most of the 13-series derivatives, showed poor absorption, indicating limited oral bioavailability. However, sorafenib showed good absorption, highlighting a potential benchmark for optimization. Despite this, none of the compounds were predicted to inhibit CYP2D6, a critical cytochrome P450 isoform involved in drug metabolism. This is a favorable attribute, suggesting low risk for drug – drug interactions and metabolic liabilities. Furthermore, all compounds demonstrated high plasma protein binding (PPB), denoted as “more than 90%,” similar to sorafenib. While high PPB may reduce free drug levels, it often contributes to prolonged plasma half-life, which can be beneficial for maintaining systemic exposure over time.

To further assess the pharmacokinetic properties of the synthesized thiadiazole derivatives, we examined two key in silico ADMET parameters: Caco-2 cell permeability (as a surrogate for intestinal absorption) and ALogP (a measure of lipophilicity).

The predicted Caco-2 permeability values for all compounds ranged from 0.5138 to 0.5648, with compound 13a showing the highest permeability (0.5648) and sorafenib, the reference drug, exhibiting the lowest (0.5138). These values, though numerically distinct, are all below the conventional high-permeability threshold (typically ≥0.9 in log scale), suggesting moderate to low intestinal absorption potential across the series. Notably, compound 11a had a permeability value of 0.5196, which is only marginally higher than that of sorafenib, reinforcing earlier observations of limited oral absorption.

Regarding ALogP, all thiadiazole derivatives exhibited higher lipophilicity than sorafenib (ALogP = 4.175). Compounds 13a and 13b recorded the highest ALogP values (6.369 and 6.191, respectively), indicating significant hydrophobicity, which may negatively impact aqueous solubility and increase plasma protein binding. Compound 11a, the most biologically promising derivative, also showed an elevated ALogP of 5.895, suggesting that its poor solubility profile may be driven by excessive lipophilicity. This trend was consistent across most analogs (Table 5).

3.3.5. In silico toxicity analysis

To further support the safety profile of the designed thiadiazole derivatives, comprehensive in silico toxicity predictions were conducted (Table S3.8.1), encompassing mutagenicity (Ames test), carcinogenicity (FDA mouse model), acute oral toxicity (LD₅₀), chronic exposure risk (LOAEL), and dermal/ocular irritancy assessments. Notably, all compounds were predicted as non-mutagenic in the Ames test, highlighting a low genotoxic potential across most of the series.

Encouragingly, none of the compounds were flagged as carcinogenic in the FDA mouse model, with the exception of the reference drug sorafenib, which was identified as a single-carcinogen. This result notably underscores a potential advantage of the newly synthesized compounds over the marketed inhibitor in terms of long-term oncogenic risk. Furthermore, the chronic oral toxicity values (LOAEL) for the novel compounds were generally low (0.035–0.061 mg/kg/day), indicating minimal chronic exposure risks, with compound 11a demonstrating a good LOAEL of 0.0611 g/kg/day. In terms of acute toxicity, represented by rat oral LD₅₀ values, compound 11a again showed a favorable profile (1.44 g/kg), indicating low acute lethality, though slightly less favorable than compounds such as 13a (1.88 g/kg). Nonetheless, all compounds displayed LD₅₀ values consistent with Class IV-V oral toxicity, supporting their potential for safe oral administration at therapeutic doses.

Dermal (DI) and ocular (OI) safety predictions further bolstered the compounds’ profiles. All thiadiazole derivatives were predicted as nonirritant to skin and only mildly irritating to the eyes, a tolerable characteristic in oral or parenteral administration contexts. These favorable irritation predictions reduce concerns of administration-site toxicity and further enhance the feasibility of systemic delivery.

Taken together, these safety data support compound 11a as not only a potent and selective anticancer agent but also one with a non-mutagenic, non-carcinogenic, and low-to-moderate toxicity profile. Its superior performance in both efficacy and safety metrics positions it as a strong candidate for further preclinical investigation and optimization.

4. Conclusion

This study introduces a structurally novel thiadiazole scaffold, offering new SAR insights and a promising basis for further development as a selective VEGFR-2–targeted anticancer agent. Among them, compound 11a demonstrated the most potent activity, with IC₅₀ values of 9.49 µM (MCF-7) and 12.89 µM (HepG-2), and a VEGFR-2 inhibition IC₅₀ of 0.055 µM, comparable to sorafenib. It showed selective cytotoxicity, induced >70% apoptosis, and caused dual-phase (S and G2/M) cell cycle arrest. Mechanistic studies confirmed activation of the intrinsic mitochondrial apoptosis pathway. Computational modeling further supports its strong and stable interaction with the VEGFR-2 active site, as well as favorable chemical reactivity and structural stability. Although in silico pharmacokinetic assessments reveal some limitations in solubility and absorption, the compound displays an encouraging safety profile, including low toxicity, non-mutagenicity, and lack of carcinogenic or irritant potential. Overall, compound 11a emerges as a well-characterized lead compound that merits further pharmacokinetics/toxicity studies and medicinal chemistry optimization to advance its development as a targeted therapeutic for VEGFR-2–driven cancers. Further studies will include testing on additional cancer cell lines and structural optimization to improve potency and selectivity. Given solubility and absorption limitations, formulation strategies such as nano-delivery or prodrugs will also be explored to enhance bioavailability.

Funding Statement

This research was funded by Princess Nourah bint Abdulrahman University Researchers Supporting Project number [PNURSP2025R116], Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia. The authors thank Research Center at AlMaarefa University for funding this work.

Article highlights

• A novel series of 2,3-dihydro-1,3,4-thiadiazole derivatives was designed and synthesized as VEGFR-2 inhibitors.

• Compound 11a showed the most potent and selective anticancer activity, with IC₅₀ values of 9.49 µM (MCF-7) and 12.89 µM (HepG-2), and a selectivity index > 3.

• VEGFR-2 enzymatic inhibition by 11a was strong (IC₅₀ = 0.055 µM), comparable to sorafenib.

• Flow cytometry confirmed that 11a induced >70% apoptosis and caused S and G2/M phase arrest.

• ELISA results indicated activation of the mitochondrial apoptotic pathway.

• Molecular docking and 200-ns MD simulations showed stable binding of 11a to the VEGFR-2 active site.

• ADMET profiling predicted low toxicity and non-mutagenicity, though solubility and permeability need improvement.

• These findings highlight compound 11a as a promising lead for further development as a selective VEGFR-2–targeted anticancer agent.

Author contributions

Eslam B. Elkaeed: Resources, Formal analysis, data collection, editing. Walid E. Elgammal: methodology. Hazem Elkady: software, writing – review & editing. Hazem A. Mahdy: writing – original draft. Aisha A. Alsfouk: funding acquisition, resources. Dalal Z. Husein: software, methodology. Omar A. Soliman: software and methodology. Mariam Omara: methodology. Ibrahim H. Eissa: project administration, supervision. Ahmed M. Metwaly: Writing – review & editing, supervision.

Disclosure statement

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.

Supplementary material

Supplemental data for this article can be accessed online at https://doi.org/10.1080/17568919.2025.2552639