Design, synthesis, docking, MD simulations, and anti-proliferative evaluation of thieno[2,3-d]pyrimidine derivatives as new EGFR inhibitors

Abstract

A group of EGFR inhibitors derived from thieno[2,3-d]pyrimidine nucleus was designed, synthesised, and examined as anti-proliferative lead compounds. MCF-7 and A549 cell lines were inhibited by 5b, the most active member. It had inhibitory partialities of 37.19 and 204.10 nM against EGFR^WT and EGFR^T790M, respectively. Compound 5b was 2.5 times safer against the WI-38 normal cell lines than erlotinib. Also, it demonstrated considerable potentialities for both early and late apoptosis induction in A549. Simultaneously, 5b arrested A549’s growth at G1 and G2/M phases. Harmoniously, 5b upregulated the BAX and downregulated the Bcl-2 genes by 3-fold and increased the BAX/Bcl-2 ratio by 8.3-fold comparing the untreated A549 cells. Molecular docking against EGFR^WT and EGFR^T790M indicated the correct binding modes. Furthermore, MD simulations confirmed the precise binding of 5b against the EGFR protein over 100 ns. Finally, various computational ADMET studies were carried out and indicated high degrees of drug-likeness and safety.

Keywords:

• Anti-proliferative
• apoptosis
• EGFR inhibitors
• MD simulations
• thieno[2,3-d]pyrimidines

Introduction

Despite the tremendous efforts to confront it, cancer remains one of the biggest public health problems facing humanity¹. According to the WHO in 2020, about 10 million deaths, or nearly one in six deaths, were caused by cancer, making it the top cause of death globally². Therefore, cancer is considered a great problem for the economy of countries as well as individuals^3,4. Oncogenic mutations, multi-drug resistance, and activation of compensatory mechanisms are examples of how the complication of cancer pathology expresses itself^5–7. Therefore, it is imperative to discover anticancer candidates that are more effective and less toxic, depending on the different biological and molecular features of cancer pathogenesis.

The protein kinases family which comprises huge types of vital enzymes and control different cellular processes including the proliferation and metastasis of human tumour cells⁸. Epidermal growth factor receptor (EGFR) is an important member of this family⁹ that has a crucial role in the growth of tumour cells¹⁰. Overexpression of EGFR is an effective prognostic factor in many types of tumours. Accordingly, hindering its molecular role exerts substantial therapeutic impacts¹¹. Thus, EGFR drew attention as an influential factor in the fight against cancer^12–14.

Erlotinib, I, is a member of the first generation of FDA-approved EGFR inhibitors exhibiting an excellent inhibitory effect against non-small-cell lung cancer¹⁵. The appearance of EGFR mutation-related resistance led scientists to discover the second generation of FDA-approved EGFR inhibitors such as neratinib II¹⁶. Unfortunately, the second generation inhibitors showed significant induced toxicity¹⁷ leading to the invention of the third generation of EGFR inhibitors such as olmutinib III^18,19 (Figure 1).

Figure 1. FDA-approved and reported EGFR inhibitors with their essential pharmacophoric features.

EGFR inhibitors share specific pharmacophoric features which are vital for the correct binding inside the ATP binding site of the EGFR protein. These essential features are represented in Figure 1 and Figure 2 including a flat hetero aromatic ring²⁰, a terminal hydrophobic head, an imino (NH, spacer) group²¹, and a hydrophobic tail^22,23. The mentioned groups are essential to occupy the adenine binding pocket, the hydrophobic region I, the space between the adenine binding region and the hydrophobic region I, and the hydrophobic region II, respectively.

Figure 2. Compound 5f as a representative example of the synthesised compounds occupying the ATP active cavity of EGFR (schematic presentation).

Thieno[2,3-d]pyrimidine is an important scaffold as it is a building block in many biologically active agents as anticancer^24,25, anti-viral²⁶, anti-oxidant²⁷, anti-inflammatory^28,29, and antimicrobial³⁰. In addition, thieno[2,3-d]pyrimidine showed good contributions in tyrosine kinase inhibitors³¹. Particularly, it was a cornerstone of many EGFR inhibitors³².

Our research team developed several 1H-pyrazolo[3,4-d]pyrimidine derivatives in 2018. The IC₅₀ results for compound IV against the wild-type EGFR^WT and mutant EGFR^T790M were very promising³³. In order to create an EGFR and HER2 tyrosine kinase inhibitor in 2019, we synthesised many thieno[2,3-d]pyrimidine derivatives. Promising action was seen in compound V³⁴. As EGFR inhibitors, we created a fresh batch of pyrimidine-5-carbonitrile derivatives in 2020. Compound VI exhibited strong inhibitory effects on the EGFR^WT and EGFR^T790M35 (Figure 1).

In this field, we used three biological isosteres (1H-pyrazolo[3,4-d]pyrimidine, thieno[2,3-d]pyrimidine, and pyrimidine-5-carbonitrile) to occupy the adenine binding pocket of EGFR. It was found that thieno[2,3-d]pyrimidine derivatives have good activities. Accordingly, we decided to modify these derivatives hoping to reach more effective anticancer agents targeting EGFR (Figure 2).

The rationale of molecular design

In the current work, the previously reported anti-EGFR compounds (IV, V, and VI) were used as lead compounds to design new EGFR inhibitors. We focussed on the thieno[2,3-d]pyrimidine nucleus to reach the new members. Such moiety was used as a hetero aromatic system to occupy the adenine binding pocket of EGFR. This moisty involves three hydrogen bond acceptors (two nitrogen atoms and a sulphur atom). This may facilitate the hydrogen bonding at the adenine binding pocket of EGFR. In addition, the fused thieno[2,3-d]pyrimidine is large enough to fill the slightly large adenine binding pocket³⁶. Furthermore, the thieno[2,3-d]pyrimidine moiety is the main nucleus of the third-generation FDA-approved EGFR inhibitor (olmutinib III) and reported EGFR inhibitor (compound V). Also, the thieno[2,3-d]pyrimidine is considered as an isostere moiety for the 1H-pyrazolo[3,4-d]pyrimidine moiety of compound IV.

The position of NH liker was kept as it is. For the hydrophobic head, we used different structures as (i) substituted phenyl moieties (compounds 5a–f), (ii) benzyl moiety (compound 6), (iii) bulky hetero-aromatic structure (compound 8), and (iv) hydrophilic moieties as thiourea group (compounds 7a) and amino group (compound 7b).

Regarding the hydrophobic tail, we used the 4-methylcyclohex-1-ene moiety in all the synthesised compounds. The 4-methylcyclohex-1-ene moiety was used to increase the hydrophobic character of the hydrophobic tail. This may increase the binding power of the new compounds in the hydrophobic pocket II (Figure 3).

Figure 3. Design strategies of the new compounds.

The synthesised compounds were evaluated for it anti-proliferative activities against two tumour cell lines. The most active members were further evaluated for their anti-EGFR activity. Deep biological investigations such as cell cycle and apoptosis analyses were conducted for the most active member. Docking studies were carried out against wild and mutant EGFR to assess the binding mode of the synthesised compounds.

Results and discussion

Chemistry

The target compounds 5a–f, 6, 7a,b, and 8 were synthesised as illustrated in Scheme 1. 4-Chloro-7-methyl-5,6,7,8-tetrahydro[1]benzothieno[2,3-d]pyrimidine 4³⁷, the key intermediate, was prepared via Gewald reaction of 4-methyl cyclohexanone 1 to give the amino ester derivative 2³⁸, which underwent cyclisation into the corresponding tricyclic 4-pyrimidione derivative 3 using formamide (99.5%)³⁷. Chlorination of the latter with POCl₃ afforded the key intermediate tricyclic 4-chloropyrimidine 4³⁷. Thereafter, 4-chloropyrimidine 4 was allowed to undergo several nucleophilic aromatic substitution reactions with aniline derivatives in acetonitrile or absolute ethanol to afford the target compounds 5a–c³⁹. Also, the reaction of the key intermediate 4 with p-phenylenediamine in isopropanol⁴⁰ afforded compound 5d. The product 5d was isolated via pouring the reaction mixture on ice. Furthermore, heating of compound 4 with benzene sulphonamide derivatives, benzyl amine, thiosemicarbazide and hydrazine hydrate (99.9%) in absolute ethanol resulted in the formation of the desired compounds 5e–f, 6, and 7a, b respectively. Finally, heating of compound 4 under reflux with p-phenylenediamine in isopropanol with stirring for 6 h afforded compound 8. The formed dimer, unexpected product 8, was obtained via filtration of the reaction mixture while hot.

Scheme 1. Synthetic pathway for the preparation of the target compounds 5a–f, 6, 7a,b, and 8.

Biology

In vitro evaluation of anti-proliferative activity

Using the conventional MTT method^41–43, all the synthesised compounds were tested for their in vitro anti-proliferative effects against a panel of two human carcinoma cell lines, the breast (MCF-7) and the lung (A549), both of which have overexpressed EGFR^44–46. The data in Table 1 demonstrated that different synthetic candidates had anti-proliferative effects on the cell lines under examination. Although compounds were generally less effective against the two tested cell lines than erlotinib, compounds 5b and 5f were the noticeably active members against A549 cells (IC₅₀ values of 17.79 and 17.46 µM) and against MCF-7 cells (IC₅₀ values of 22.66 and 21.40 µM) comparing erlotinib (IC₅₀ values of 4.18 and 14.27 µM), respectively. On the other side, the rest of the candidates, except compound 8, were moderately active expressing IC₅₀ values ranging from 18.39 to 35.68 µM.

Table 1. In vitro anti-proliferative and EGFR inhibitory activities of the tested compounds.

Samples	Cell lines IC50 (µM)^a			Enzymes IC50 (nM)^a
	A549	MCF-7	WI-38	EGFR^WT	EGFR^T790M
5a	20.61	29.32	NT^b	NT^b	NT^b
5b	17.79	22.66	70.13	37.19	204.10
5c	27.14	30.04	NT^b	NT^b	NT^b
5d	34.77	35.48	NT^b	NT^b	NT^b
5e	19.90	29.23	NT^b	NT^b	NT^b
5f	17.46	21.40	NT^b	35.64	NT^b
6	21.15	26.55	NT^b	NT^b	NT^b
7a	28.69	35.68	NT^b	NT^b	NT^b
7b	18.39	23.70	NT^b	NT^b	NT^b
8	43.06	56.19	NT^b	NT^b	NT^b
Erlotinib	4.18	14.27	28.48	5.90	212.20

^aData are presented as the mean of the IC₅₀ values from three different experiments.

^b NT = Not tested.

Structure-activity relationships

Comparing the cytotoxicity of compounds 5a–f (incorporating substituted phenyl moieties), compound 6 (incorporating a benzyl moiety), and compound 8 (incorporating a bulky hetero-aromatic structure), revealed that the substituted phenyl moieties are more advantageous than the benzyl moiety, and the latter is more preferred than the bulky hetero-aromatic structure.

Regarding the substituted phenyl moieties (compounds 5a–f), it was found that the substitution with electron-withdrawing groups (compounds 5a, 5b, 5e, and 5f) is more advantageous than the substitution with electron-donating ones (compounds 5c and 5d). For the substitution with electron-withdrawing groups at the 4-position of the phenyl ring (compounds 5a and 5f), it was found the SO₂NH₂ group is more active than the chloro atom. For the substitution with electron-withdrawing groups at the 2-position of the phenyl ring (compounds 5b and 5e), it was found the fluoro atom is more active than the SO₂NH₂ group. Regarding the substitution with electron-donating groups at the 4-position of the phenyl ring (compounds 5c and 5d), it was found the OCH₃ group is more active than the NH₂ group. Finally, comparing the activities of compounds 7a (incorporating thiourea moiety) and 7b (incorporating amino group), it was found that the amino group is more advantageous than the thiourea moiety.

Toxicity of compound 5b towards normal human cells

The unintended toxicity caused to healthy cells by the anticancer drugs is one of the key serious side effects of chemotherapy. Also, it’s essential to examine if the obtained anti-proliferative activities of 5b originated from its anticancer potentiality or general toxicity. Accordingly, normal human lung fibroblasts, WI-38, cell lines were utilised to evaluate the safety of 5b. Table 1’s findings regarding 5b’s in vitro anti-proliferative activities demonstrate that it appeared to be non-toxic to WI-38 with an IC₅₀ of 70.13 µM compared to that of erlotinib (28.48 µM).

Evaluation of kinases activity

For the active candidate, 5b, against EGFR^WT and EGFR^T790M, a homogeneous time-resolved fluorescence (HTRF) assay⁴⁷ was used to assess its inhibitory activities. As a reference standard, erlotinib was used. With IC₅₀ values of 37.19 and 204.10 nM against EGFR^WT and mutant EGFR^T790M, compound 5b had comparable inhibitory actions to those of erlotinib (IC₅₀ = 5.9 and 212.2 nM, respectively) (Table 1).

Cell cycle analysis

Eukaryotic cells can replicate themselves through the successive cell cycle. Any anti-proliferative drug affects the targeted cells and stops their growth at one or more checkpoints (phases) of growth. If cellular damage persists, checkpoint signalling may trigger several systems that start apoptosis. Finding such a likely stage of cell arrest is essential for the drug discovery process⁴⁸.

In this work, A549 cells were subjected to 5b at a concentration of 17.79 µM for 72h. Investigating the changes that occurred in the cell cycle, it was noticed that 5b induced a significant decrease in cells’ population at the S phase (37.66%) and at the Sub-G1 phase (0.54%), comparing to the untreated A549 cells (71.62% and 1.14%, respectively). Simultaneously, 5b exerted a substantial increase in the A549 population at the G1 phase (23.19%), and at the G2/M phase (38.61%), comparing to the untreated A549 cells (9.12% and 18.12%, respectively) (Figure 4). This indicated that compound 5b can arrest the A549 cells at both G1 and G2/M phases (Table 2).

Figure 4. Flow cytometry analysis of A549 cell cycle phases post treatment with 5b.

Table 2. Effect of 5b on cell cycle progression in A549 cells after 72 h treatment.

Sample	Cell cycle distribution (%)^a
	%Sub-G1	%G1	%S	% G2/M
A549	1.14 ± 0.12	9.12 ± 1.34	71.62 ± 1.64	18.12 ± 2.86
5b/A549	0.54 ± 0.09	23.19 ± 2.18*	37.66 ± 4.19*	38.61 ± 2.11*

^aValues are given as mean ± SEM of two independent experiments.

*p < 0.05 indicates statistically significant differences from the corresponding control (A549) group in unpaired t-tests.

Apoptosis analysis

The apoptotic effect of 5b, 17.79 µM, on A549 cells was evaluated using the Annexin-V/propidium iodide staining (Annexin-V FTIC) test. As shown in Table 3 and Figure 5, compound 5b significantly increased and induced apoptosis (early and late) in the A549 cells to be 17.84% and 45.71%, respectively comparing the untreated A549 cells (0.07% and 0.98%, respectively.

Figure 5. Flow cytometry analysis of apoptosis in A549 cells exposed to compound 5b.

Table 3. Effect of compound 5b on stages of the cell death process in A549 cells after 72 h treatment.

Sample	Viable^a (left bottom)	Apoptosis^a		Necrosis^a (left top)
		Early (right bottom)	Late (right top)
A549	98.93 ± 0.04	0.07 ± 0.03	0.98 ± 0.03	0.02 ± 0.01
5b/ A549	34.33 ± 2.20	17.84 ± 3.68	45.71 ± 5.77**	2.12 ± 0.11

^aValues are given as mean ± SEM of two independent experiments.

**p < 0.01 indicates a statistically significant difference from the corresponding control (A549) group in unpaired t-tests.

Effects on the mitochondrial apoptosis pathway

BAX and Bcl-2 are two primary regulating genes controlling mitochondrial apoptosis. During the apoptotic cascade, these mediators serve opposing functions. In contrast with Bcl-2 (anti-apoptotic), BAX exhibits a pro-apoptotic activity [43]. Therefore, the BAX/Bcl-2 ratio is a key factor to understand mitochondrial apoptosis [44]. In this test, A549 cells were treated with compound 5b at a concentration of 17.79 µM for 72h. The investigation of BAX and Bcl-2 genes expression by quantitative RT-PCR (Figure 6 and Table 4) showed the significant effects of 5b (17.79 µM) against the A549 cells. Interestingly, 5b upregulated the BAX and downregulated the Bcl-2 genes by 3-fold each. Simultaneously, 5b increased the BAX/Bcl-2 ratio by 8.3 times comparing the untreated A549 cells.

Figure 6. Gene expression analysis of BAX and Bcl-2 the expression levels after treatment of A549 with compound 5b for 72h.

Table 4. Effect of 5b on levels of BAX and Bcl-2 genes expression in A549 cells treated for 72 h.

Sample	Gene expression (fold change)^a
	BAX	Bcl-2	BAX/Bcl-2 ratio
A549	1.00 ± 0.19	1.00 ± 0.17	1.00 ± 0.03
5b/A549	2.97 ± 0.14**	0.37 ± 0.06*	8.35 ± 1.09**

^aValues are given as changes from the corresponding control (A549) group, which is set to “1”.

*p < 0.05 **p < 0.01 indicate statistically significant differences from the corresponding control in unpaired t-tests.

In silico studies

Docking studies

The docking studies were carried out against two forms of EGFR (Wild-type (EGFR^WT, PDB: 4HJO)⁴⁹ and mutant type (EGFR^T790M, PDB: 3W2O)⁵⁰) to assess the binding efficiency of the synthesised compounds against wild and mutant types of the EGFR protein. The co-crystallised ligands (erlotinib and TAK-285) were used as reference molecules. MOE 2019 software was used in these studies. The binding scores of the synthesised compounds and the co-crystallised ligands were summarised in Table 5.

Table 5. The docking binding free energies of the synthesised compounds against EGFR^WT and EGFR^T790M.

Comp.	Binding free energy (kcal/mol)
	EGFR^WT	EGFR^T790M
5a	−17.54	−16.01
5b	−17.22	−16.08
5c	−18.64	−17.27
5d	−17.41	−16.27
6	−18.81	−17.61
7a	−15.02	−12.95
5e	−19.84	−16.11
5f	−17.49	−17.37
7b	−12.87	−12.29
8	−26.27	−21.08
Erlotinib	−23.94	–
TAK-285	–	−21.55

The binding pattern of erlotinib was similar to the reported results³⁵. It showed a binding score of −23.94 kcal/mol. Two hydrogen bonds (HBs) and seven hydrophobic interactions (HIs) stabilised the molecule in the active site. Each function group occupied a particular pocket in the active site as appeared in Figure 7.

Figure 7. Erlotinib docked into the active site of EGFR^WT forming two HBs with Met769 and Cys773 and seven HIs with Lys721, Val702, Ala719, Leu820, and Leu694.

Compound 5b exhibited a similar binding pattern to erlotinib. It showed a binding score of −17.22 kcal/mol. It formed one HBs with Met769. In addition, it formed twelve HIs with Lys721, Val702, Ala719, Leu820, Cys773, and Leu694. The different moieties of compound 5b occupied the active pockets of EGFR. The thieno[2,3-d]pyrimidine, fluorobenzene, and 4-methylcyclohex-1-ene moieties occupied the adenine pocket, the hydrophobic pocket I, and the hydrophobic pocket II, respectively (Figure 8).

Figure 8. Compound 5b docked into the active site of EGFR^WT forming 1 HB with Met769 and 12 HIs with Lys721, Val702, Ala719, Leu820, Cys773, and Leu694.

Compound 5f exhibited showed a binding score of −17.49 kcal/mol with a similar binding mode to erlotinib. It formed an HB with Met769 and ten HIs with Val702, Ala719, Leu820, Cys773, and Leu694. The thieno[2,3-d]pyrimidine, benzenesulphonamide, and 4-methylcyclohex-1-ene moieties occupied the adenine pocket, the hydrophobic pocket I, and the hydrophobic pocket II, respectively (Figure 9).

Figure 9. Compound 5f docked into the active site of EGFR^WT, forming 1 HB with Met769, and 10 HIs with Val702, Ala719, Leu820, Cys773, and Leu694.

The binding modes of the tested compounds against EGFR^T790M were similar to that of the co-crystallised ligand (TAK-285). The latter showed a binding score of −21.55 kcal/mol. It formed three HBs with Met793, Ser720, and Lys745. In addition, it formed fourteen HIs with Lys745, Glu762, Leu788, Ile759, Leu844, Ala743, Val726, Met790, and Ala743. The pyrrolo[3,2-d]pyrimidine, 3-(trifluoromethyl)phenoxy, N-ethyl-3-hydroxy-3-methylbutanamide moieties were buried in the adenine pocket, hydrophobic pocket I, and hydrophobic region II, respectively (Figure 10).

Figure 10. Co-crystallised ligand (TAK-285) docked into the active site of EGFR^T790M forming formed 3 HBs with Met793, Ser720, and Lys745 and 14 HIs with Lys745, Glu762, Leu788, Ile759, Leu844, Ala743, Val726, Met790, and Ala743.

Compound 5f exhibited showed a binding score of −17.37 kcal/mol with a similar binding mode to TAK-285. It formed two HBs with Met793 and Lys745. In addition, it formed nine HBs with Met766, Leu788, Ile759, Val726, Met790, Ala763, and Ala743. Furthermore, it formed three electrostatic attractions with Met790, Lys745, and Glu762. The thieno[2,3-d]pyrimidine, benzenesulphonamide, and 4-methylcyclohex-1-ene moieties occupied the adenine pocket, the hydrophobic pocket I, and the hydrophobic pocket II, respectively (Figure 11).

Figure 11. Binding of compound 5f with EGFRT^790M, forming two HBs with Met793 and Lys745, nine HIs with Met766, Leu788, Ile759, Val726, Met790, Ala763, and Ala743, and three electrostatic attractions with Met790, Lys745, and Glu762.

Molecular dynamic simulations

The EGFR protein shows consistent behaviour with a steady fluctuation of the 5b molecule, according to the analysis done for the EGFR-5b complex throughout the production run. Figure 12(A) (blue and green curves) shows that the EGFR and the EGFR-5b complex RMSD plots stabilise after 32 ns, with an average value of 2.11 Å for both of them for the rest of the trajectory. Furthermore, the RMSD of the 5b demonstrates a consistent trend around 2.69 Å throughout the trajectory. The RoG (Figure 12(B)) exhibits a very little rising trend, reaching 19.7 Å at the end of the simulation with a change of only 0.4 Å from the initial frame. The SASA (Figure 12(C)) shows a consistent average of 15 188 Ų. Figure 12(D) depicts a decrease in the number of HBs throughout the first 32 ns until it stabilises at an average of 53 HBs which indicates the stability of the protein conformation. Except for Ser696:Phe699, Ala840:Lys851, and the free C-terminal, which reach 2.2 Å, 3.3 Å, and 6.6 Å, respectively, the variation of the amino acids represented in the RMSF plot (Figure 12(E)) is relatively low (less than 2 Å). The 5b has a constant separation of 11.3 Å between its centre of mass and the protein’s centre of mass (Figure 12(F)).

Figure 12. (A) RMSD values from the trajectory for the EGFR protein (blue curve), 5b (red curve), and EGFR & 5b complex (green curve), (B) radius of gyration, (C) SASA, (D) change in the number of HBs, (E) RMSF, and (F) distance from the centre of mass of 5b and EGFR protein.

MM-GBSA analysis

The different components that contribute to the MM-GBSA binding free energy analysis are depicted in Figure 13. The overall binding of the 5b is −23.94 Kcal/Mol on average. The van der Waals energy has the most beneficial contribution with an average value of −37.31 Kcal/Mol, followed by the electrostatic interaction energy with an average value of −9.53 Kcal/Mol. Furthermore, we performed decomposition analysis (Figure 14) to determine which amino acids within 1 nm of the 5b molecule contribute to the interaction with a value less than −1 kcal/mol. These amino acids are Leu694 (−1.56 Kcal/Mol), Val702 (−1.4 Kcal/Mol), Ala719 (−1.07 Kcal/Mol), Thr766 (−1.0 Kcal/Mol), Leu768 (−1.39 Kcal/Mol), Met769 (−1.68 Kcal/Mol), Cys773 (−1.15 Kcal/Mol), and Leu820 (−1.75 Kcal/Mol). Moreover, two amino acids have positive free energy contributions to the binding (Asp776 (+0.97 Kcal/Mol) and Asp831 (+1.39 Kcal/Mol)).

Figure 13. MM-GBSA analysis of the EGFR-5b complex.

Figure 14. Binding free energy decomposition of the EGFR-5b complex.

PLIP analysis

The trajectory of the EGFR-5b complex was clustered, and then a representative frame for each cluster was generated. The elbow method was used to determine the number of clusters, and four clusters were generated. The PLIP website was used to detect the number and type of interactions between the 5b and the EGFR protein for each cluster representative. Table 6 displays the number and types of interactions retrieved from the PLIP website. Two types of interactions have been reported from the PLIP webserver with a large difference in their numbers. Only one amino acid (Met769) is forming H-bond in all representatives. On the other hand, there are 20 HIs detected with Leu694, Thr766, and Lys721 as the most common amino acids. In addition to producing the interaction types and numbers from PLIP, it also generates a .pse file to see the 3D conformation of the ligand and its interaction with the protein (Figure 15).

Figure 15. PLIP analysis of the EGFR-5b complex for each cluster representative. HB: blue solid line, HI: dashed grey line, amino acids: blue sticks, and 5b: orange sticks.

Table 6. The number and types of interactions detected from the PLIP webserver.

Cluster number	No. of HBS	Amino acids in EGFR	No. of HIS	Amino acids in EGFR
C1	1	Met769	5	Leu694–Val702–Lys721–Thr766–Leu820
C2	1	Met769	4	Leu694–Ala719–Lys721–Thr766
C3	1	Met769	6	Leu694–Val702–Ala719–Lys721–Leu764–Thr766
C4	1	Met769	5	Leu694–Val702–Lys721–Thr766–Leu820

Note: Bold amino acids are the amino acids with the largest number of interactions in all cluster representatives.

ADMET analysis

Since the approval of a novel medicine is based on its pharmacokinetic assessment and its biological activity, it is important to explore the pharmacokinetic qualities of a new chemical early on in the drug discovery process to prevent delays in approval or retraction⁵¹. Using erlotinib as a reference molecule, the ADMET characteristics of the synthesised compounds were evaluated computationally using Discovery Studio 4.0 (Figure 16 and Table 7). Although most compounds were predicted to have bad levels of aqueous solubility (Sol. L), most of them expressed good levels of intestinal absorption (Abs. L). Most of the tested compounds were anticipated to have high or very high levels of BBB penetration levels. Regarding the cytochrome P4502D6 inhibitory effects, most compounds were predicted as non-inhibitors. The level of plasma protein binding (PPB. L) model predicts that most compounds were expected to bind plasma protein over 90%.

Figure 16. The expected ADMET study.

Table 7. Calculated ADMET descriptors.

Comp.	BBB	Sol. L	Abs. L	CYP2D6	PPB. L
5a	Very high	Very low	Good	Maybe an inhibitor	>90%
5b
5c	High			Non-inhibitor
5d
6	Very high			Maybe an inhibitor
7a	Medium	Low		Non-inhibitor	<90%
5e	Very low	Very low			>90%
5f
7b	Medium	Low			< 90%
8	Very low	Very low	Very Poor		>90%
Erlotinib	High	Low	Good

Toxicity profiles analysis

For a successful drug development to be minimised the estimation of toxicity is of the utmost importance early on in the process⁵¹. In addition, the use of in silico approaches played an indispensable role in the development of drugs to avoid ethical regulations and wasting resources and time on the usual in vitro ad in vivo studies⁵². An in silico toxicity prediction, or SAR-predictive toxicity, compares the basic chemical structural properties of molecules being tested with the structures of thousands of compounds with reported safety and toxicity⁵³.

According to the toxicity models built in Discovery Studio software, seven models of toxicity were estimated computationally. These models are; median carcinogenic dose TD₅₀ (MC- TD₅₀) in rats, Maximum Tolerated Dose, in Rats by Feeding (MTDRF), Developmental Toxicity Potential (DTP), Rat Oral LD₅₀ (RLD₅₀), Chronic Lowest-Observed-Adverse-Effect Level (CLOAEL) in addition to the potentialities to be irritant against eye (OI) and skin (SI). As shown in Table 8, most compounds showed in silico low toxicity profile against the tested models. Interestingly, compound 5b was predicted to have safer patterns than that of erlotinib in two models (MC-TD₅₀ and MTDRF). Simultaneously, compound 5b was predicted to have almost similar safety patterns to that of erlotinib in four models (CLOAEL, DTP OI, and SI). On the other hand, it was less safe in the model of RLD_50.

Table 8. Toxicity properties of the synthesised compounds.

Comp.	MC-TD50^a	MTDRF^b	DTP	RLD50^b	CLOAEL^b	OI	SI
5a	3.7461	0.129531	Non-toxic	0.483059	0.0375019	Irritant	Non-irritant
5b	10.4772	0.138151		0.211542	0.0746137
5c	1.41938	0.0606517		0.450393	0.048643
5d	2.41863	0.137552		0.610853	0.0630163
6	29.8656	0.100448		0.359547	0.0709221		Irritant
7a	10.0095	0.221518		0.169431	0.12173		Non-irritant
5e	29.2894	0.108657		2.94856	0.323061
5f	10.6751	0.108657		2.61225	0.177435
7b	9.14814	0.133206		0.466603	0.129244
8	0.347134	0.0912244		0.705064	0.0718845
Erlotinib	8.05746	0.0827884		0.662169	0.0359487

^aUnit: mg/kg/day.

^bUnit: g/kg.

Conclusion

As anti-proliferative lead compounds, a series of thieno[2,3-d] pyrimidine-derived EGFR inhibitors was designed, synthesised, and examined. The most active member, 5b inhibited MCF-7, and A549 cell lines (IC₅₀ = 22.66, and 17.79 µM, respectively). The inhibitory partialities of 5b against two isoforms: EGFR^WT and EGFR^T790M were assessed to be 37.19, and 204.10 nM, respectively. Interestingly, 5b was much safer (2.5-fold) against the normal cell lines, WI-38, (IC₅₀ =70.13 µM) than erlotinib (IC₅₀ =28.48 µM). Regarding the mechanism of 5b’s anticancer activities, it showed significant potentialities to induce early and late apoptosis in A549 cells based on annexin V staining and arrested the G1 and G2/M phases of A549 cells. To better comprehend and explain the obtained activities on computational levels, molecular docking, MD simulations, MM-GPSA, and PLIP studies against the EGFR protein were performed and confirmed the accurate binding. Therefore, this study presents compound 5b as a potential lead anticancer and EGFR inhibitor.

Experimental

Chemistry

All materials and apparatus used in the synthesis and analyses of the synthesised compounds were clarified in Supplementary Material. Compounds 2 and 3 were synthesised according to the reported procedures^54,55.

4-Chloro-7-methyl-5,6,7,8-tetrahydro[1]benzothieno[2,3-d]pyrimidine (4)

A mixture of thienopyrimidone 3 (1.1 g, 5 mmol), phosphorus oxychloride (15 ml) and pyridine (1 ml) was heated under reflux for 5 h. The oily residue was poured portion wise into ice cold water (100 ml), and the formed solid was dried and crystallised from ethanol.

Yellow solid (0.9 g, 80%). MP = 100–102 °C; IR (KBr, cm⁻¹): 3059 (CH aromatic), 2954 (CH aliphatic), 1561 (C = N), 1492 (C = C) cm⁻¹; ¹HNMR (400 MHz, DMSO-d₆, δ ppm) δ 8.79 (s, 1H, C-₂H), 3.13–3.12 (m, 2H, CH₂), 2.69–2.64 (m, 2H, CH₂), 2.21–2.14 (m, 2H, CH₂), 1.45–1.42 (m, 1H, CH), 1.06–1.00 (m, 3H, CH₃); ¹³C NMR (101 MHz, DMSO-d₆, δ ppm) δ 152.77 (C-₂), 152.25 (C-Cl), 144.45, 139.64, 131.19, 128.25 (ArCs), 30.17, 29.21, 28.67, 25.99 (4 aliphatic Cs), 21.41 (CH₃).

General procedure for the synthesis of 7-methyl-N-(substituted-phenyl)-5,6,7,8-tetrahydro [1]benzothieno[2,3-d] pyrimidin-4-amine (5a–c)

To a solution of compound 4 (0.24 g, 1 mmol) in acetonitrile for compound 5b or absolute ethanol for compounds 5a and 5c (20 ml), appropriate aniline derivatives (1.1 mmol) were added and the reaction mixture was heated under reflux for 24 h. The formed precipitate was collected, dried and crystallised from ethanol to give the target compounds 5a–c.

N-(4-Chlorophenyl)-7-methyl-5,6,7,8-tetrahydrobenzo[4,5]thieno[2,3-d]pyrimidin-4-amine (5a)

Buff solid (0.3 g, 85%). MP = 170–172 °C; IR (KBr, cm⁻¹): 3450 (NH), 3112 (CH aromatic), 2927 (CH aliphatic), 1662 (C = N), 1563, 1439 (C = C) cm⁻¹; ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 8.40 (s, 1H, C₂-H), 8.23 (s, 1H, NH, D₂O exchangeable), 7.71 (d, J = 8.4 Hz, 2H, 4-Cl-C₆H₄-C_2,6-H), 7.40 (d, J = 8.4 Hz, 2H, 4-Cl-C₆H₄-C_3,5-H), 3.21–3.18 (m, 2H, CH₂), 2.93–2.88 (m, 1H, CH), 2.46–2.39 (m, 1H, CH), 1.94 (d, J = 8 Hz, 2H, CH₂), 1.46 (s, 1H, CH), 1.08 (d, J = 8 Hz, 3H, CH₃); ¹³C NMR (101 MHz, DMSO-d₆, δ ppm) δ 166.55 (C-₂), 155.04 (C-₄), 152.42 (C-Cl), 138.77 (C-NH), 133.25, 128.70, 127.33, 126.67, 124.01, 117.27 (ArCs), 33.44, 30.48, 28.82, 25.46 (4 aliphatic Cs), 21.48 (CH₃). APCI-MS m/z: Exact mass calcd for C₁₇H₁₆ClN₃S [M]⁺: 329.1. Found: 329.9.

N-(2-Fluorophenyl)-7-methyl-5,6,7,8-tetrahydrobenzo[4,5]thieno[2,3-d]pyrimidin-4-amine (5b)

Reddish brown solid (0.2 g, 70%). MP = 125–127 °C; IR (KBr, cm⁻¹): 3455 (NH), 3090 (CH aromatic), 2951, 2927 (CH aliphatic), 1623 (C = N), 1506, 1450 (C = C) cm⁻¹; ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 8.34 (s, 1H, C₂-H), 8.12 (s, 1H, NH, D₂O exchangeable), 7.84 (t, J = 8 Hz, 1H, 2-F-C₆H₄-C₄-H), 7.30 (d, J = 8 Hz, 1H, 2-F-C₆H₄-C₃-H), 7.26 (d, J = 4 Hz, 1H, 2-F-C₆H₄-C₆-H), 7.22 (t, J = 8 Hz, 1H, 2-F-C₆H₄-C₅-H), 3.21–3.17 (m, 1H, CH), 3.09–3.07 (m, 1H, CH), 2.93–2.88 (m, 1H, CH), 2.46–2.40 (m, 1H, CH), 1.97–1.94 (m, 2H, CH₂), 1.54–1.44 (m, 1H, CH), 1.08 (d, J = 8 Hz, 3H, CH₃), ¹³C NMR (101 MHz, DMSO-d₆, δ ppm) δ 166.32 (C-₂), 157.46 (C-₄), 155.55 (C-F), 155.02 (C-NH), 152.67, 133.14, 127.47, 126.85, 126.59, 124.79, 116.95, 115.87 (ArCs), 33.39, 30.47, 28.80, 25.64 (4 aliphatic Cs), 21.48 (CH₃). APCI-MS m/z: Exact mass calcd for C₁₇H₁₆FN₃S [M]⁺: 313.1. Found: 313.8.

N-(4-Methoxyphenyl)-7-methyl-5,6,7,8-tetrahydrobenzo[4,5]thieno[2,3-d]pyrimidin -4-amine (5c)

Violet solid (0.2 g, 62%). MP = 140–142 °C; IR (KBr, cm⁻¹): 3450 (NH), 3101 (CH aromatic), 2951, 2835 (CH aliphatic), 1606 (C = C), 1564, 1443 (C = N) cm⁻¹. ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 8.64 (s, 1H, C₂-H), 8.44 (s, 1H, NH, D₂O exchangeable), 7.49 (d, J = 8 Hz, 2H, 4-OCH₃-C₆H₄-C_3,5-H), 6.99 (d, J = 8 Hz, 2H, 4-OCH₃-C₆H₄-C_2,6-H), 3.78 (s, 3H, OCH₃), 3.23–3.14 (m, 2H, CH₂), 2.96–2.92 (m, 1H, CH), 2.48–2.42 (m, 1H, CH), 1.94 (s, 2H, CH₂), 1.51–1.47 (s, 1H, CH), 1.09 (d, J = 8 Hz, 3H, CH₃), ¹³C NMR (101 MHz, DMSO-d₆, δ ppm) δ 157.24 (C-₂), 155.58 (C-₄), 150.17 (C-OCH₃), 133.76 (C-NH), 127.26, 126.27, 124.98, 116.74, 115.29, 114.34 (ArCs), 55.81 (OCH₃), 33.32, 30.33, 28.74, 25.49 (4 aliphatic Cs), 21.44 (CH₃). APCI-MS m/z: Exact mass calcd for C₁₇H₁₇N₃OS [M]⁺: 311.1. Found: 314.1.

N¹-(7-Methyl-5,6,7,8-tetrahydro[1]benzothieno[2,3-d]pyrimidin-4-yl)benzene-1,4-diamine (5d)

4-Chloro-7-methyl-5,6,7,8-tetrahydro[1]benzothieno[2,3-d]pyrimidine (4) (0.24 g, 1 mmol) was added to a solution of p-phenylenediamine (0.108 g, 1 mmol) in isopropanol (25 ml). The reaction mixture was stirred at 100 °C for 3 h. After cooling, the reaction mixture was poured onto crushed ice with continuous stirring. The formed solid was filtered off, washed with water then dried and crystallised from isopropanol. Gray solid (0.2 g, 65%). MP = 170–172 °C; IR (KBr, cm⁻¹): 3421, 3351 (NH₂, NH), 3119 (CH aromatic). 2947, 2922 (CH aliphatic), 1624 (C = N), 1559, 1481 (C = C) cm⁻¹; ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 8.27 (s, 1H, C₂-H), 7.85 (s,1H, NH, D₂O exchangeable), 7.26 (d, J = 7.9 Hz, 2H, 4-NH₂-C₆H₄-C₃,₅-H), 6.65 (d, J = 7.8 Hz, 2H, 4-NH₂-C₆H₄-C₂,₆-H), 5.79 (s, 2H, NH₂, D₂O exchangeable), 3.18–3.09 (m, 2H, CH₂), 2.93–2.86 (s, 1H, CH), 2.45–2.41 (m, 1H, CH), 1.95 (s, 2H, CH₂), 1.47 (s, 1H, CH), 1.10 (d, J = 10 Hz, 3H, CH₃), ¹³C NMR (101 MHz, DMSO-d₆, δ ppm) δ 165.72 (C-₂), 156.16 (C-₄), 152.80, 145.82 (C-NH), 131.92 (C-NH₂), 128.18, 126.77, 125.53, 116.28, 114.26 (ArCs), 33.41, 30.58, 28.82, 25.70 (4 aliphatic Cs), 21.51 (CH₃). APCI-MS m/z: Exact mass calcd for C₁₇H₁₈N₄S [M]⁺: 310.1. Found: 311.2.

General procedure for the synthesis of N-substituted-((7-methyl-5,6,7,8-tetrahydro [1]benzothieno[2,3-d]pyrimidin-4-yl) amino)benzenesulphonamide (5e,f)

Chloropyrimidine 4 (0.24 g, 1 mmol) and the appropriate substituted benzenesulphonamide (1 mmol) in isopropanol (25 ml) was heated under reflux for 10 h. The reaction mixture was allowed to cool, the solid formed was filtered, dried and crystallised from absolute ethanol to give compounds 5e,f.

2-((7-Methyl-5,6,7,8-tetrahydrobenzo[4,5]thieno[2,3-d]pyrimidin-4-yl)amino)benzene sulphonamide (5e)

Buff solid (0.3 g, 83%). MP = 200–202 °C; IR (KBr, cm⁻¹): 3482, 3391, 3251 (NH₂, NH), 3052 (CH aromatic), 2925 (CH aliphatic), 1617 (C = N), 1532, 1414 (C = C) cm⁻¹; ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 8.81 (s, 1H, C₂-H), 7.54 (d, J = 8 Hz, 1H, 2-H₂NO₂S-C₆H₄-C-₃-H), 7.23 (t, J = 8 Hz 1H, 2-H₂NO₂S-C₆H₄-C-₄-H), 7.11 (s, 1H, NH, D₂O exchangeable), 6.79 (d, J = 8 Hz, 1H, 2-H₂NO₂S-C₆H₄-C-₆-H), 6.60 (t, J = 8 Hz, 1H, 2-H₂NO₂S-C₆H₄-C-₅-H), 5.60 (s, 2H, NH₂, D₂O exchangeable), 3.21–3.17 (m, 2H, CH₂), 3.02–2.92 (m. 2H, CH₂), 1.96 (d, J = 8 Hz, 2H, CH₂), 1.51–1.45 (m, 1H, CH), 1.08 (d, J = 8 Hz, 3H, CH₃). ¹³C NMR (101 MHz, DMSO-d_6, δ ppm) δ 168.66 (C-₂), 157.07 (C-₄), 152.82, 152.29, 146.02 (C-NH), 139.70, 133.32 (C-SO₂-NH₂), 128.34, 126.81, 124.68, 117.15, 115.42, (ArCs), 33.68, 30.18, 28.68, 25.99 (4 aliphatic Cs), 21.41 (CH₃).

4-((7-Methyl-5,6,7,8-tetrahydrobenzo[4,5]thieno[2,3-d]pyrimidin-4-yl)amino)benzene sulphonamide (5f)

Buff solid (0.3 g, 79%). MP = 210–212 °C; IR (KBr, cm⁻¹): 3436, 3325, 3200 (NH₂, NH), 3040 (CH aromatic), 2956 (CH aliphatic), 1604 (C = N), 1562, 1439 (C = C) cm⁻¹; ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 8.51 (s, 1H, C₂-H), 8.48 (s, 1H, NH, D₂O exchangeable), 7.85 (d, J = 8.8 Hz, 2H, 4-H₂NO₂S-C₆H₄-C-_2,6-H), 7.79 (d, J = 8.8 Hz, 2H, 4-H₂NO₂S-C₆H₄-C-_3,5-H), 7.28 (s, 2H, NH₂, D₂O exchangeable), 3.25–3.18 (m, 2H, CH₂), 2.97–2.92 (m, 1H, CH), 2.47–2.43 (m, 1H, CH), 1.98–1.95 (m, 2H, CH₂), 1.54–1.44 (m, 1H, CH), 1.10 (d, J = 8 Hz, 3H, CH₃); ¹³C NMR (101 MHz, DMSO-d₆, δ ppm) δ 166.93 (C-₂), 161.61 (C-₄), 154.74, 152.37, 143.01 (C-NH), 138.34, 133.84 (C-SO₂-NH₂), 126.69, 121.24, 117.87 (ArCs), 33.47, 30.45, 28.83, 25.36 (4 aliphatic Cs), 21.47 (CH₃).

N-((7-Methyl-5,6,7,8-tetrahydro[1]benzothieno[2,3-d]pyrimidin-4-yl)methyl) aniline (6)

A solution of chloropyrimidine 4 (0.24 g, 1 mmol) in absolute ethanol (25 ml) was treated with benzylamine (0.12 g, 1.1 mmol) and heated under reflux for 24 h. The solid obtained after cooling was dried and crystallised from isopropanol.

Buff solid (0.2 g, 64%) MP = 140–142 °C; IR (KBr, cm⁻¹): 3403 (NH), 3052 (CH aromatic), 2933, 2852 (CH aliphatic), 1610 (C = N), 1510, 1445 (C = C) cm⁻¹; ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 8.81 (s, 1H, C₂-H), 8.25 (s, 5H, ArH), 5.87 (s, 1H, NH, D₂O exchangeable, 4.09 (s, 2H, CH₂-NH), 3.01–2.97 (m, 1H, CH), 2.85–2.81 (m, 1H, CH), 2.39–2.33 (m, 1H, CH), 1.72 (d, J = 8 Hz, 1H, CH), 1.61 (d, J = 8 Hz, 2H, CH₂), 1.23–1.17 (m, 1H, CH), 1.05 (d, J = 4 Hz, 3H, CH₃). ¹³C NMR (101 MHz, DMSO-d₆, δ ppm) δ 165.13 (C-₂), 156.68 (C-₄), 153.20, 152.27, 139.65 (C-NH), 131.29, 128.31, 126.61, 115.69 (ArCs), 49.46 (CH₂-NH), 33.26, 30.48, 28.73, 25.19, (4 aliphatic Cs), 21.45 (CH₃). APCI-MS m/z: Exact mass calcd for C₁₈H₁₉N₃S [M]⁺: 309.1. Found: 309.1.

2–(7-Methyl-5,6,7,8-tetrahydro[1]benzothieno[2,3-d]pyrimidin-4-yl)hydrazine carbothioamide (7a)

Chloropyrimidine 4 (0.24 g, 1 mmol) and thiosemicarbazide (0.1 g, 1 mmol) was dissolved in absolute ethanol (25 ml) and heated under reflux for 8 h. The precipitated solid was filtered, dried and crystallised from ethanol.

Orange solid (0.3 g, 82%). MP = 213–215 °C; IR (KBr, cm⁻¹): 3494, 3428, 3301 (NH₂, two NHs), 3077 (CH aromatic), 2944 (CH aliphatic), 1647 (C = N), 1536, 1460 (C = C) cm⁻¹; ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 9.17 (s, 1H, C₂-H), 8.44 (s, 1H, NH, D₂O exchangeable), 8.10 (s, 1H, NH, D₂O exchangeable), 6.49 (s, 2H, NH₂, D₂O exchangeable), 3.17–3.13 (m, 1H, CH), 2.97–2.88 (m, 2H, CH₂), 2.46–2.42 (m, 1H, CH), 1.98–1.89 (m, 2H, CH₂), 1.49–1.48 (m, 1H, CH), 1.09 (d, J = 6.6 Hz, 3H, CH₃); ¹³C NMR (101 MHz, DMSO-d₆, δ ppm) δ 178.51 (C = S), 167.16 (C-₂), 153.55 (C-₄), 148.71 (C-NH), 136.42, 135.89, 128.08 (ArCs), 33.20, 30.39, 29.31, 25.19 (4 aliphatic Cs), 21.56 (CH₃). APCI-MS m/z: Exact mass calcd for C₁₂H₁₅N₅S₂ [M]⁺: 293.08.

4-Hydrazinyl-7-methyl-5,6,7,8-tetrahydro[1]benzothieno[2,3-d]pyrimidine (7b)

Hydrazine hydrate (0.1 g, 2 mmol, 99%) solution in absolute ethanol (25 ml) was added to chloropyrimidine 4 (0.24 g, 1 mmol) and heated under reflux for 5 h. After cooling, the mixture was poured on ice-cooled water, filtered, dried and crystallised from ethanol to give the hydrazinyl derivative 7b.

Buff solid (0.2 g, 58%). MP = 180–182 °C; IR (KBr, cm⁻¹): 3421, 3346 (NH₂, NH), 3082 (CH aromatic), 2922, 2861 (CH aliphatic), 1636 (C = N), 1560, 1473 (C = C) cm⁻¹; ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 8.33 (s, 1H, C₂-H), 7.87 (s, 1H, NH, D₂O exchangeable), 4.65 (s, 2H, NH₂, D₂O exchangeable), 3.03–2.81 (m, 2H, CH₂), 2.40–2.34 (m, 2H, CH₂), 1.90–1.87 (m, 2H, CH₂), 1.43–1.39 (m, 1H, CH), 1.05 (d, J = 8 Hz, 3H, CH₃); ¹³C NMR (101 MHz, DMSO-d_6, δ ppm) δ 164.49 (C-₄), 158.68 (C-₂), 152.93, 131.56, 126.95, 115.17 (ArCs), 33.30, 30.60, 28.85, 25.61 (4 aliphatic Cs), 21.51 (CH₃).

N¹,N-4-Bis(7-methyl-5,6,7,8-tetrahydrobenzo[4,5]thieno[2,3-d]pyrimidin-4-yl) benzene-1,4-diamine (8)

A solution of chloropyrimidine 4 (0.24 g, 1 mmol) and p-phenylenediamine (0.12 g, 1.1 mmol) in isopropanol (25 ml) was heated under reflux for 6 h with stirring. The mixture was filtered on hot, washed with ethanol and dried to give the dimer 8.

Buff solid (0.3 g, 55%). MP = 220–222 °C; IR (KBr, cm⁻¹): 3439, 3400 (two NHs), 3020 (CH aromatic), 2949, 2918 (CH aliphatic), 1585 (C = N), 1443 (C = C) cm⁻¹; ¹H NMR (400 MHz, DMSO-d₆, δ ppm) δ 8.36 (s, 2H, two C₂-H), 8.11 (s, 2H, two NHs, D₂O exchangeable), 7.61 (s, 4H, ArH), 3.25–3.15 (m, 4H, two CH₂), 2.94–2.90 (m, 2H, two CH), 2.51–2.41 (m, 2H, two CH), 1.98–1.95 (m, 4H, two CH₂), 1.51–1.49 (m, 2H, two CH), 1.09 (d, J = 4 Hz, 6H, two CH₃); ¹³C NMR (101 MHz, DMSO-d₆, δ ppm) δ 174.44 (C-₄), 165.85 (C-₂), 157.76, 152.75, 132.17, 126.77, 125.20, 123.19, 116.42, (ArCs), 33.41, 30.53, 28.81, 25 0.95 (4 aliphatic Cs), 21.48 (CH₃); APCI-MS m/z: Exact mass calcd for C₂₈H₂₈N₆S₂ [M]⁺: 512.1. Found: 511.1.

Biological examinations

In vitro anti-proliferative activity

The anti-proliferative activities of the synthesised compounds against A549 and MCF-7cell lines were carried out using an MTT procedure as shown in Supplementary Material^42,56,57.

In vitro EGFR inhibition

The inhibitory activities of the synthesised compounds against EGFR^WT and EGFR^T790M proteins were carried out using EGFR Kinase Assay Kit (BPS Bioscience, USA) as shown in Supplementary Material⁴⁷.

Safety assay

The cytotoxic effect of compound 5b against a normal cell line (W138) was assessed using an MTT assay.

Cell cycle analysis

The cell cycle analysis was carried out using a flow cytometer as shown in Supplementary Material^58,59.

Apoptosis analysis

The apoptotic effect of compound 5b was tested using a flow cytometer as shown in Supplementary Material^60,61.

BAX and Bcl-2 determination

Supplementary Material provides a thorough justification. The effect of compound 5b on BAX and Bcl-2 level in A549 cells^62–64.

In silico studies

Docking studies

Docking studies against EGFR^WT and EGFR^T790M were performed by MOE2019 software as shown in Supplementary Material³⁵.

MD simulations

MD simulation studies were directed by CHARMM-GUI web server and GROMACS 2021 as an MD engine as shown in Supplementary Material^65,66.

MM-GBSA and PLIP

MM-GBSA and PLIP analyses were performed by the Gmx_MMPBSA package as shown in Supplementary Material⁶⁷.

ADMET studies

ADMET profile was examined by Discovery Studio 4.0 as shown in Supplementary Material⁶⁸.

Toxicity studies

The toxicity profile was tested by Discovery Studio 4.0 as shown in Supplementary Material⁶⁹.

Acknowledgement

The authors extend their appreciation to the Research Center at AlMaarefa University for funding this work.

Disclosure statement

The authors report no conflicts of interest.

Additional information

Funding

This research was funded by Princess Nourah bint Abdulrahman University Researchers Supporting Project number (PNURSP2023R116), Princess Nourah bint Abdulrahman University, Riyadh, Saudi Arabia. The authors extend their appreciation to the Research Center at AlMaarefa University for funding this work.