Design, synthesis and molecular docking of new fused 1H-pyrroles, pyrrolo[3,2-d]pyrimidines and pyrrolo[3,2-e][1, 4]diazepine derivatives as potent EGFR/CDK2 inhibitors

Abstract

A new series of 1H-pyrrole (6a–c, 8a–c), pyrrolo[3,2-d]pyrimidines (9a–c) and pyrrolo[3,2-e][1, 4]diazepines (11a–c) were designed and synthesised. These compounds were designed to have the essential pharmacophoric features of EGFR Inhibitors, they have shown anticancer activities against HCT116, MCF-7 and Hep3B cancer cells with IC₅₀ values ranging from 0.009 to 2.195 µM. IC₅₀ value of doxorubicin is 0.008 µM, compounds 9a and 9c showed IC₅₀ values of 0.011 and 0.009 µM respectively against HCT-116 cells. Compound 8b exerted broad-spectrum activity against all tested cell lines with an IC₅₀ value less than 0.05 µM. Compound 8b was evaluated against a panel of kinases. This compound potently inhibited CDK2/Cyclin A1, DYRK3 and GSK3 alpha kinases with 10–23% compared to imatinib (1–10%). It has also arrested the cell cycle of MCF-7 cells at the S phase. Its antiproliferative activity was further augmented by molecular docking into the active sites of EGFR and CDK2 cyclin A1.

Keywords:

• Anticancer
• 1H-pyrrole
• pyrrolo[3,2-d]pyrimidine
• pyrrolo[3,2-e][1
• 4]diazepine
• EGFR inhibitor
• CDK2 inhibitor

1. Introduction

Cancer is clearly associated with an increased incidence and rates of mortality, in addition to its devastating social and economic effects^1,2. Cancer is a generic term for a large group of diseases that can affect any part of the human body^2,3. The complexity of cancer pathologies manifests in oncogenic mutations, severe and occasionally fatal drug side effects, multi-drug resistance and activation of compensatory pathways^4–6. Thus, there is an urgent need to develop more efficient anticancer candidates, relying on various biological and molecular aspects of neoplastic transformation. Accordingly, many efforts were done to reach promising anticancer candidates^1,7–10.

Protein kinases constitute one of the largest and most functionally diverse gene families that control a diverse set of cellular processes. Thus, they perform a major role in the proliferation, metastasis and survival of human tumour cells¹¹. Tyrosine kinase is the most important protein kinases sub-family. This family comprises epidermal growth factor receptor (EGFR)¹², vascular endothelial growth factor receptor (VEGFR)¹³, fibroblast growth factor receptor (FGFR)¹⁴ and cyclin-dependent kinases (CDK)¹⁵. Three of the protein kinases have been crystallised in active conformations (cAPK, PhK, and CK1), one in a partially active conformation (CDK2 in complex with cyclin A), and five in inactive conformations (CDK2, MAPK, IRK, twitchin kinase, and CaMKI). The structures of the kinases in the active conformation all showed equivalent positions for essential catalytic site residues Lys-72, Asp-166, and Asp-184 in cAPK. Their positions and the correct orientation of the Mg²⁺/ATP and the protein substrate appear crucial for catalysis and are dependent upon the tertiary structure of both lobes and the correct relative orientation of the lobes. In the partially active CDK2–cyclin A complex, the charge grouping is compensated by glutamate, Glu-162, (two residues removed from the phosphorylatable threonine Thr-160) and by interactions of arginyl residues with the main chain carbonyl groups from cyclin A. In PhK, there is only one basic group, the arginine adjacent to the catalytic base, and charge compensation can be satisfactorily accomplished by a carboxylate group¹⁶.

The epidermal growth factor receptor (EGFR) is a transmembrane protein receptor endowed with tyrosine kinase activity, occupying an important position in cancer progression and tumour cell biology. EGFR plays a role in cellular phenotyping and provides tumour cells with significant growth¹⁷. Elevated levels of EGFR and its associated ligands (EGF and transforming growth factor (TGF)) have been found to be a common feature of a variety of cancer types. In many cases, abnormal EGFR activation appears to be a key element in carcinogenesis and a major driving force for cancer growth¹⁸. Increased EGFR expression is thus expected to be a powerful prognostic factor in a variety of tumour forms and blocking its cellular functions looks to have significant therapeutic effects¹⁹. As a result, EGFR is being regarded as a viable target for the development of novel anticancer drugs^20–23.

The cyclin-dependent kinases (CDKs) are a group of enzymes involved in cell cycle progression and cellular proliferation²⁴. They work by phosphorylating critical serine and threonine residues in host proteins, which then can activate them^25,26. It is commonly believed that inhibiting CDKs could help the limitation of the uncontrolled cellular proliferation seen in some malignancies²⁷. The majority of CDK inhibitors bind to the ATP pocket as ATP-competitive inhibitors with essential structural hydrogen-bonding motifs²⁸. Cyclin-dependent kinase 2 (CDK2) belongs to the serine/threonine protein kinase family, and the CDK2 activity is found to be typically high in different types of human cancers. Studies have reported that CDK2 overexpression indicates poor prognosis in patients with HCC, and inhibition of CDK2 activity could reverse the malignant phenotype of cancer cells. Many studies revealed a key role of CDK2 in EGF-induced cell transformation and the associated signal transduction pathways²⁹. The literature survey revealed that EGFR and CDK2 were the key targets for many antitumor agents e.g. cinobufagin³⁰. In addition, biological and computational evidence supported that anticancer agents such as benzamide-substituted chalcones exerted their anti-proliferative effects via dual EGFR/CDK2 inhibitory activities³¹.

It is worth mentioning that, dual-specificity tyrosine phosphorylation-regulated kinase 3 (DYRK3) is a regulator of phase transition during mitosis³². Furthermore, it is able to promote mTORC1 activity, which is associated with resistance to EGFR-mediated endocrine therapy, and other forms of targeted therapy, also it regulates fundamental cellular functions including transcription, translation, proliferation, growth and survival³³. Moreover, Glycogen synthase kinase 3 (GSK3) is involved in modulating numerous signalling pathways affecting metabolism, tumorigenesis, proliferation, apoptosis, autophagy, development, and differentiation involved in metabolic regulation³⁴. GSK3 inhibitors were reported to suppress breast tumour growth in pre-clinical models^35,36.

Until now, there are many FDA-approved drugs for the treatment of cancer targeting EGFR as erlotinib I and afatinib II. The first generation of these inhibitors (as erlotinib I) gave good benefits in the treatment of non-small-cell lung cancer³⁷. The second generation of EGFR inhibitors (as afatinib II) was approved to overcome EGFR mutation-related resistance³⁸. The third generation of EGFR inhibitors (as rociletinib III) had overcome induced toxicity caused by 2^nd generation^39–41. Palbociclib IV is the first FDA-approved cyclin-dependent kinase inhibitor that acts by binding to the ATP pocket with an IC₅₀ in the range of 9–15 nM²⁷ (Figure 1).

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

EGFR inhibitors have some pharmacophoric features which are essential for maximal affinity against the ATP binding site of EGFR. These features include i) a flat hetero aromatic ring system that occupies the adenine binding pocket⁴², ii) terminal hydrophobic head which can occupy the hydrophobic region I, iii) imino group (NH, spacer) which can occupy the space between the adenine binding region and the hydrophobic region I⁴³, iv) hydrophobic tail which occupies the hydrophobic region II^44,45, ribose binding moiety which can occupy the ribose binding pocket. Till now, there are limitations in research that target the ribose binding pocket⁴⁶ (Figure 2).

Figure 2. ATP binding site of EGFR cavity composed of five main parts^47–49.

Privileged structures of diazepine have been widely used as effective templates for anticancer agents especially targeting EGFR enzyme⁵⁰. Xu et al., described in 2013 a compound with a 1,3‐diazepine moiety (compound V) presenting a potent inhibition activity against the EGFR790M mutant (IC₅₀ = 3.36 nM) and the double EGFRL858R/T790M mutant (IC₅₀ = 1.69 nM)⁵¹. Also, in 2019, a series of thieno[2,3-d]pyrimidine derivatives were designed and synthesised as EGFR and HER2 tyrosine kinase inhibitors. Compound VI exhibited promising cytotoxic activity⁴⁷. In 2020, a new series of pyrimidine-5-carbonitrile derivatives were designed and synthesised as EGFR inhibitors. Compound VII showed high inhibitory activities against EGFR^WT and EGFR^T790M48 (Figure 1).

Based on the mentioned facts and as an extension of the previous efforts in the design and synthesis of new anticancer agents^48,52–59 especially to target EGFR^47,48,60 we used the 1H-pyrrole, pyrrolo[3,2-d]pyrimidine, and pyrrolo[3,2-e][1, 4]diazepine moieties as building blocks for the design and synthesis of new EGFR and CDK inhibitors (Figure 3).

Figure 3. Design strategies of Scaffolds A, B and C.

1.1. Rationale of molecular design

In the present work, the previously active compounds (V, VI and VII) were used as lead compounds to design new EGFR inhibitors. Dramatic modifications were achieved to reach more promising active candidates against EGFR and CDK. The modifications were performed on four features. The flat hetero aromatic system was modified to be 1H-pyrrole (compounds 6a–c and 8a–c) (Scaffold A), pyrrolo[3,2-d]pyrimidine (9a–c) (Scaffold B), and pyrrolo[3,2-e][1, 4]diazepine (compounds 11a–c) (Scaffold C). ii) For the terminal hydrophobic head, we used different hydrophobic moieties as substituted phenyl groups (compounds 6a–c) or aliphatic moieties (compounds 8a–c, 9a–c, and 11a–c). Regarding the hydrophobic tail, we used the pyrrolidine moiety in all the designed compounds. The NH spacer was kept as it is in compounds 8a–c, whereas it was changed into -N = moiety in compounds 6a–c, and deleted in compounds 9a–c relaying on the two nitrogen atoms of pyrimidine moiety to act as hydrogen bond centres. Moreover, in compounds 11a–c, it was modified to be inside the ring structure as a heteroatom. To occupy ribose binding moieties, we used different aromatic structures. Different structure modifications were then achieved to obtain a SAR study as a second aim in our work.

To confirm our rationale, the synthesised compounds were investigated for their antiproliferative activities against a panel of human cancer cell lines (Hep3B, HCT116, and MCF-7). In addition, the most active anti-proliferative agent was further subjected to a Kinase profiling test to assess its activity against EGFR, CDK, and other kinases. Furthermore, to reach a good insight into the activity of the most active candidate at the molecular level, cell cycle analysis was carried out. Finally, in silico, docking, ADMET, and toxicity studies were performed to predict the possible binding interaction of the synthesised compounds against the prospective targets (EGFR and CDK) as well as to calculate the proposed kinetic and toxicity profile.

2. Results and discussion

2.1. Chemistry

Compounds 8⁶¹ and 10a–c⁶² were synthesised according to reported methods, which were then used to afford the intermediates 11a–c according to the reported procedure⁶³. Condensation of compounds 5a–c (Scheme 1) with 4-nitrobenzaldehyde in absolute ethanol afforded three Schiff base derivatives (6a–c, Scheme 2). The IR spectra of compounds 6a–c showed sharp absorption bands at 2210–2214 cm⁻¹_, indicating the cyano group. Furthermore, the ¹H NMR spectra of compounds 6a–c showed two doublet signals at a range of 8.10–8.42 ppm due to the aromatic protons of the 4-nitrobenzylidene moiety. Moreover, a singlet signal appeared at a range of 9.28–9.31 corresponding to the N = CH proton. While their ¹³C NMR spectra revealed two signals at a range of 156.3–158 ppm indicating the N=CH and the carbonyl carbon in each of the three derivatives. The free amino group of compounds 5a–c was further acylated with 2-chloroacetyl chloride in benzene to give compounds 7a–c. Compounds 8a–c were prepared by N-alkylation of piperidine with the terminal chlorinated side chain of 7a–c in absolute ethanol, using sodium bicarbonate to neutralise the hydrogen chloride liberated from this reaction (Scheme 2). The ¹H NMR spectra of compounds 8a–c confirmed the aliphatic protons of the piperidin-1-yl ring; multiplet and triplet signals at a range of 1.50–2.74 ppm. Their ¹³C NMR spectra further revealed three additional signals in the range of 23.4–55.0 ppm due to the aliphatic carbons of the piperidin-1-yl ring.

Scheme 1. Construction of compounds 5a–c. Reagents and conditions: (a) (CH₃)₂SO₄, benzene, CH₂(CN)₂, reflux, 6 h; (b) ClCH₂COCl, glacial acetic acid, CH₂COONa, 30–40 °C, 2 h; (c) acetone, K₂CO₃, reflux, 24 h.

Scheme 2. Synthesis of compounds 6a–c, 7a–c and 8a–c. Reagents and conditions: (a) 4-Nitrobenzaldehyde, absolute ethanol, glacial acetic acid; (b) ClCH₂COCl, benzene; r.t.; 48 h; (c) piperidine, NaHCO₃, absolute ethanol, reflux, 8 h.

Moreover, several studies reported the use of triethyl orthoformate to produce cyclized derivatives⁶⁴. The tricyclic pyrrolo[3,2-d]pyrimidine derivatives 9a–c were achieved by refluxing compounds 5a–c with excess triethyl orthoformate (Scheme 3). The NHs stretching bands disappeared in the IR spectra of compounds 9a–c. Further, the ¹H NMR spectra of compounds 9a–c confirmed disappearance of the exchangeable protons’ signals, corresponding to compounds 5a–c NHs protons. Additionally, they showed singlet signals at a range of 2.25–2.27 ppm corresponding to the aliphatic methyl protons. Furthermore, the ¹³C NMR spectra revealed the presence of two additional signals; the first one at the range of 24.3–24.4 ppm due to the aliphatic methyl group at C-2 and a second signal at the range of 153.7–154.1 ppm assigned for the C-2. Whereas, the pyrrolo[3,2-e][1, 4]diazepine derivatives 10a–c were obtained by intramolecular cyclisation of compounds 7a–c using potassium carbonate in dimethylformamide (Scheme 3). Finally, N-alkylation of compounds 10a–c with ethyl chloroacetate afforded the ethyl ester derivatives 11a–c using potassium carbonate in acetone (Scheme 3). The disappearance of the amidic protons (CONH) of the starting materials 10a–c was observed in the ¹H NMR spectra of compounds 11a–c, in addition to the presence of triplet signals at 1.28–1.29 ppm and quartette signals around 4.24 ppm due to the ethyl group. The ¹³C NMR spectra of compounds 11a–c revealed three signals in the range of 159.6–167.6 ppm that were assigned for the carbon atoms of the three carbonyl groups.

Scheme 3. Synthesis of compounds 9a–c, 10a–c and 11a–c. Reagents and conditions: (a) Excess CH₃C(OC₂H₅)₃, reflux, 12 h; (b) K₂CO₃, DMF, r.t., 48 h; (c) ClCH₂COOC₂H₅, K₂CO₃, acetone, reflux, 6 h.

2.2. Biological evaluation

2.2.1. Anticancer activity

Evaluation of the anticancer activity was performed against liver (Hep3B), colon (HCT116), and breast (MCF-7) cancer cell lines in the Centre of Genetic Engineering, Al-Azhar University, Cairo, Egypt, using Sulforhodamine-B (SRB) assay⁶⁵. Doxorubicin (DOX) was used as a reference drug. The survival curve was obtained by plotting concentrations of the compound under investigation against the survival fraction of the tumour cells. Then, results were expressed in half maximal inhibitory concentration (IC₅₀).

All the tested derivatives showed potent antiproliferative activities (IC₅₀ = 0.009 − 2.195 µM) against the three cancer cell lines (Table 1). A closer look at the results revealed that compound 8b exhibited the highest cytotoxic activity against Hep3B and MCF-7 cell lines with IC₅₀ values of 0.049 and 0.043 µM, respectively. On the other hand, compound 8b displayed the highest cytotoxic activity against the HCT116 cell line with IC₅₀ values of 0.011 µM.

Table 1. IC₅₀ values of the new compounds (6a–c, 8a–c, 9a–c, and 11a–c) against Hep3B, HCT116 and MCF-7 cell lines.

Comp.	Scaffold	R	Spacer	R1	IC50 (µM)
					Hep3B	HCT116	MCF-7
6a	A	H	N = CH-		0.219	0.213	0.422
6b	A	CH3	N = CH-		1.956	2.069	0.336
6c	A	Cl	N = CH-		1.470	0.955	2.195
8a	A	H	-NHCOCH2-		0.313	0.408	0.315
8b	A	CH3	-NHCOCH2-		0.049	0.031	0.043
8c	A	Cl	-NHCOCH2-		0.144	0.180	0.294
9a	B	H	–	–	0.487	0.011	0.137
9b	B	CH3	–	–	0.265	0.820	0.644
9c	B	Cl	–	–	0.072	0.009	1.479
11a	C	H	–	–	0.218	0.840	0.154
11b	C	CH3	–	–	0.271	0.303	0.354
11c	C	Cl	–	–	0.179	0.286	0.364
DOX	-	-	-	–	0.005	0.008	0.008

Hep3B, Liver cancer cell line; HCT116, colon cancer cell line; MCF-7, breast cancer cell line; DOX, doxorubicin.

2.2.2.1. Structure-Activity Relationship (SAR)

Inspecting the results of the anti-proliferative activity of the designed candidates, we concluded a valuable SAR. With reference to their cytotoxic activity, it was noticed that counterparts incorporating 2,3-dihydro-1H-pyrrolizine moiety (scaffold A) were slightly more advantageous than the 3H-pyrimido[4,5-b]pyrrolizine derivatives (scaffold B). However, hexahydro-[1, 4]diazepino[5,6-b]pyrrolizin containing derivatives (scaffold C) displayed less potent inhibitory activity against the tested cell lines.

Concerning the di-aryl compounds 6a–c, compound 6a (unsubstituted-N-phenylcarbamoyl derivatives) displayed higher activity than 6b and 6c (4-methyl and 4-chloro-N-phenylcarbamoyl derivatives, respectively) against Hep3B (IC₅₀ = 0.219 µM) and HCT116 (IC₅₀ = 0.213 µM) cancer cell lines. While the methyl analogue 6b was the most active member against MCF-7 cell line with IC₅₀ value of 0.336 µM. Replacement of the aromatic 6–4-nitrobenzylidene)amino side chain in compounds 6a–c by the 2-(piperidin-1-yl)acetamido moiety (compounds 8a–c) resulted in a slightly enhanced cytotoxicity; compounds 8a–c showed IC₅₀ values in the range of 0.031–0.408 µM against the three cell lines. Among the three derivatives 8a–c, the methyl derivative 8b displayed the highest anticancer activity against the three cell lines.

Evaluation of the anticancer activity of pyrrolo[3,2-d]pyrimidine 9a–c, revealed that the 4-chloro congener derivative 9c was the most active against HCT116 cell line among all the tested compounds (IC₅₀ = 0.009 µM). While, the pyrrolo[3,2-e][1, 4]diazepine derivatives (compounds 11a–c) showed comparable activity to their fused pyrrolo[3,2-d]pyrimidine analogues (9a–c) with enhancement in activity of the 4-chloro 11c over 9c against MCF-7 cell line (IC₅₀ = 0.364 versus 1.479 µM). Thus, the SAR study pointed out the significance of a single methyl group as it presents in the most active compounds 8b (N-p-tolylcarbamoyl derivative) and 9a,c; 2-methyl group on the fused pyrrolo[3,2-d]pyrimidine scaffold. However, the comparable activity among the three series (A–C scaffolds) is raising a query about the significance of the molecular modifications on the physicochemical properties, which could affect the cellular barriers positively.

2.2.2. EGFR and CDK-2 inhibitory activity

According to the anti-proliferative activity of the tested compounds, 8b showed a substantial broad-spectrum activity against Hep3B, HCT116 and MCF-7 cell lines (IC₅₀ = 0.049, 0.031 and 0.043 µM, respectively). Thus, compound 8b was tested for its inhibitory activity against both EGFR and CDK-2.

The results revealed that compound 8b showed excellent activity against CDK2 with an inhibition value of 15% compared to the reference molecule, imatinib (2%). On the other hand, it showed less activity against EGFR (% Inhibition = −70%) compared to imatinib (% Inhibition = −13%) (Table 2).

Table 2. Inhibitory activity of compound 8b and imatinib against 20 kinases at 10 µM⁶⁶.

Kinase	% Inhibition		Kinase	% Inhibition
	Imatinib^a	8b^b		Imatinib^a	8b^b
AMPK (A1/B1/G1)	−14%	−18%	EPHA1	−22%	−46%
ALK1	−18%	−21%	FLT1	10%	−8%
ASK1	−5%	−23%	GRK1	−2%	−1%
Aurora A	−38%	−6%	GSK3 alpha	1%	10%
BLK	−3%	−36%	MSK1	−49%	−48%
BRAF	−13%	−77%	NEK1	2%	−8%
CDK2/Cyclin A1	2%	15%	p38 Alpha	−17%	−23%
CK1 Alpha 1	0%	−10%	PDK1	−34%	−28%
DYRK3	10%	23%	PRKG1	−4%	−4%
EGFR	−13%	−70%	SGK1	−3%	−10%

The + ve values indicate the inhibition % and − ve values indicate increase in enzyme activity.

2.2.3. Kinase profiling test

The previous results encouraged us to test compound 8b against other different kinases (18 kinases) to reach a good insight into its kinases inhibitory profile. The result showed that compound 8b has good inhibitory activity against DYRK3 and GSK3 alpha kinases. For its activity against DYRK3, it showed an inhibition value of 23% compared to imatinib (% Inhibition = 10%). Regarding its activity against GSK3 alpha, it showed an inhibition value of 10% compared to imatinib (% Inhibition = 1%).

Additionally, compound 8b showed imatinib comparable activity against ALK1, AMPK (A1/B1/G1), GRK1, MSK1, p38 Alpha, PDK1, and PRKG1. On the other hand, it showed moderate to weak activity against CK1 Alpha 1, Aurora A, BLK, BRAF, ASK1, EPHA1, FLT1, NEK1, and SGK1 (Table 2).

2.2.4. Determination of cell cycle perturbations

To study the mechanistic actions regarding compounds 6a and 8b, cell cycle analyses using flow cytometry (BC, FC500) were performed. The results of the cell cycle perturbation of the MCF-7 cell line treated with compounds 6a and 8b (72 h, separately) were presented in Figures 4 and 5, respectively. Each of the two compounds caused a three-fold increase of cells in the S phase at 1 µM compared to control. This increase in the S-phase cell population was accompanied by a concomitant decrease in the G₁ cell population. However, there was no considerable change in the S phase cellular population at 5 and 10 µM, which means that the effect of compounds 6a and 8b are not dose-dependent.

Figure 4. Cell cycle distribution of MCF-7 treated with compound 6a (µM, 72 h: x axis); % cell (y axis).

Figure 5. Cell cycle distribution of MCF-7 treated with compound 8b (µM, 72 h: x axis); % cell (y axis).

2.3. In silico studies

2.3.1. Docking studies

A docking study was conducted in the hopes of learning more about how the synthesised compounds interact with their targets^67–69. EGFR (PDB: 4HJO), CDK-2 (PDB: 6GUH), DYRK3 (PDB: 5Y86), and GSK3 (PDB: 5HLP) were employed as biological targets in docking investigations utilising MOE 14.0 software. The co-crystallised ligands (erlotinib and AZD5438) were utilised as anti-EGFR and anti-CDK-2 reference compounds respectively. When compared to the reference molecules, the docking results demonstrated. that the synthesised compounds have a high affinity for the two examined targets (Table 3).

Table 3. The binding free energies of the synthesised compounds against EGFR and CDK-2.

Comp.	Binding free energy (kcal/mol)
	EGFR	CDK-2
6a	−18.33	−23.67
6b	−18.97	−23.84
6c	−19.14	−23.60
8a	−20.67	−24.89
8b	−22.46	−24.52
8c	−20.08	−24.20
9a	−17.69	−17.77
9b	−18.90	−17.74
9c	−17.72	−16.71
11a	−21.54	−24.08
11b	−23.30	−24.18
11c	−18.93	−23.98
Erlotinib	−23.49	–
AZD5438	–	−21.66

Redocking of co-crystallised ligands (Erlotinib and AZD5438) against EGFR and CDK-2, respectively, was used to validate docking experiments. Erlotinib and AZD5438 had RMSDs of 0.88 and 0.54 Å for docked and original ligands, respectively. The validity of the docking operation is shown by these values (Figure 6).

Figure 6. (A) superimposition of the docked ligand of erlotinib (turquoise) and the original ligand (green) with an RMSD value of 0.88 Å. (B) superimposition of the docked ligand of AZD5438 (pink) and the original ligand (green) with RMSD value of 0.54 Å.

The binding energy of erlotinib was found to be −23.49 kcal/mol. The quinazoline molecule was buried in the adenine pocket, creating a hydrogen connection with Met769. Furthermore, Lue694, Ala719, and Leu820 established four hydrophobic contacts with quinazoline. The ethynylphenyl moiety was positioned in the hydrophobic pocket I, resulting in three hydrophobic interactions with Ala719, Val702, and Lys721. The 2-methoxyethoxy groups formed a hydrogen bond with Cys773 in the hydrophobic region II (Figure 7).

Figure 7. Erlotinib docked into the active site of EGFR.

Compound 8b showed a binding mode like that of erlotinib with a binding energy of −22.46 kcal/mol. The 1H-pyrrole-3-carbonitrile moiety occupied the adenine pocket of the EGFR forming three hydrophobic interactions with Leu820, Leu694, and Val702. Piperidine moiety was oriented into the hydrophobic pocket I to form one hydrophobic interaction with Leu694. The pyrrolidine moiety occupied the hydrophobic pocket II forming one hydrophobic interaction with Lys721 in close contact with Thr766, Leu764, and Leu834. Moreover, p-tolyl moiety occupied the ribose binding pocket forming one electrostatic attraction with Cys773 (Figure 8).

Figure 8. Compound 8b docked into the active site of EGFR.

The co-crystallised ligand (AZD5438) showed binding energy of −21.66 kcal/mol against CDK-2. The pyrimidine moiety occupied the adenine pocket forming three hydrophobic bonds with Leu134, and Ala31, Val64. The NH-linker formed one hydrogen bond with Leu83. The terminal methylsulfonyl benzene moiety occupied the hydrophobic pocket forming a hydrogen bond with Asp86 and one hydrophobic interaction with Ile10. Additionally, the 1-isopropyl-2-methyl-1H-imidazole moiety occupied the other hydrophobic region forming a hydrogen bond with Lys33. Also, it formed four hydrophobic interactions with Leu134 and Val18 (Figure 9).

Figure 9. Co-crystallised ligand (AZD5438) docked into the active site of CDK-2.

Compound 8b showed a binding mode like that of AZD543 with a binding energy of −24.52 kcal/mol. The 1H-pyrrole-3-carbonitrile moiety occupied the adenine pocket forming two hydrophobic interactions with Leu134 and Val18. The pyrrolidine moiety was oriented into the hydrophobic pocket forming three hydrophobic interactions with Val18, Lys33, and Ala31. The piperidine moiety occupied another hydrophobic pocket forming one hydrophobic interaction with Lys89 in close contact with Asp86 and Gln85. The two amide linkers formed one hydrogen bond and one electrostatic interaction with Ile10. Moreover, p-tolyl moiety occupied the ribose binding pocket forming one hydrophobic interaction with Gly13 (Figure 10).

Figure 10. (A) Binding of compound 8b with CDK-2.

Docking calculations between compound 8b and DYRK3 and GSK3 revealed that it could occupy the active sites and interact with the key amino acids of each enzyme (Figure 11 and Figure 12) with the binding energy of −24.54 and −16.85 kcal/mol, respectively. The binding free energy of the co-crystallised ligands of DYRK3 (HRM) and GSK3 (65 A) was calculated to be −20.80 and −16.41 kcal/mol, respectively and the binding modes were presented in Supplementary data.

Figure 11. (A) Binding of compound 8b with DYRK3.

Figure 12. (A) Binding of compound 8b with GSK3 alpha.

2.3.2. In silico toxicity prediction

In this work, eight toxicity parameters were estimated computationally depending on the constructed toxicity models in Discovery studio software^70–72. The results revealed that the calculated toxicity potential of the synthesised compound was low. Except for compound 6a, all compounds were non-carcinogenic for mouse females based on the FDA rodent carcinogenicity model. In addition, compounds 6a, 8a, 8b, 9a, and 11a showed carcinogenic potency TD₅₀ values ranging from 11.774 to 71.541 mg/kg body weight/day, which were higher than that of erlotinib (8.057 mg/kg body weight/day). All compounds had rat maximum tolerated dose values less than that of erlotinib (0.083 g/kg body weight) except compound 8c (0.107 g/kg body weight). For the rat oral LD₅₀ model, except compounds 9a and 9c, all compounds showed oral LD₅₀ values higher than that of erlotinib (0.662 mg/kg body weight/day). In addition, all compounds exhibited rat chronic LOAEL values higher than that of erlotinib (0.036 mg/kg body weight/day) except compounds 8c and 11c. Moreover, all the tested compounds were predicted to be mild irritants in the ocular irritancy model, and non-irritants in the skin irritancy model (Table 4).

Table 4. Toxicity properties of the synthesised compounds.

Comp.	FDA rodent carcinogenicity (mouse-female)	Carcinogenic potency TD50 (Rat)^a	Rat maximum tolerated dose (feed)^b	Rat oral LD50^b	Rat chronic LOAEL^b	Ocular irritancy	Skin irritancy
6a	Single-Carcinogen	18.143	0.060	0.800	0.157	Irritant	Non-Irritant
6b	Non-Carcinogen	2.356	0.049	2.050	0.096	Irritant	Non-Irritant
6c	Non-Carcinogen	1.839	0.073	0.838	0.063	Irritant	Non-Irritant
8a	Non-Carcinogen	71.541	0.087	1.162	0.071	Irritant	Non-Irritant
8b	Non-Carcinogen	11.813	0.072	3.451	0.054	Irritant	Non-Irritant
8c	Non-Carcinogen	7.264	0.107	1.219	0.029	Irritant	Non-Irritant
9a	Non-Carcinogen	11.774	0.045	0.433	0.163	Irritant	Non-Irritant
9b	Non-Carcinogen	3.458	0.037	0.829	0.137	Irritant	Non-Irritant
9c	Non-Carcinogen	2.160	0.056	0.531	0.073	Irritant	Non-Irritant
11a	Non-Carcinogen	22.882	0.025	1.344	0.080	Irritant	Non-Irritant
11b	Non-Carcinogen	6.639	0.021	2.522	0.067	Irritant	Non-Irritant
11c	Non-Carcinogen	4.081	0.031	1.603	0.035	Irritant	Non-Irritant
Erlotinib	Non-Carcinogen	8.057	0.083	0.662	0.036	Irritant	Non-Irritant

^aUnit: mg/kg body weight/day.

^bUnit: g/kg body weight.

2.3.3. ADMET studies

ADMET parameters were predicted using discovery studio software^69,73–75. Erlotinib was used as a reference compound. The predicted ADMET parameters were listed in Table 5. The results revealed that compounds 6a–c had very low Blood Brain Barrier penetration power. Compounds 8a–c and 11a–c were anticipated to have low levels of BBB penetration. On the other hand, compound 9a–c was predicted to have medium levels of BBB penetration. Accordingly, the synthesised compounds were expected to be safe for CNS. The predicted aqueous solubility of the synthesised compounds ranged from good to low. All the synthesised compounds showed good absorption levels except compounds 6a and 6b which showed moderate absorption levels. All the synthesised members were predicted as non-inhibitors of CYP2D6. Additionally, all of them were expected to bind plasma protein by more than 90% (Figure 13).

Figure 13. The expected ADMET study.

Table 5. Predicted ADMET profile for the synthesised compounds

Comp.	BBB level^a	Solubility level^b	Absorption level^c	CYP2D6 prediction^d	PPB prediction^e
6a	4	2	1	False	True
6b	4	2	1	False	True
6c	4	2	2	False	True
8a	3	3	0	False	True
8b	3	2	0	False	True
8c	3	2	0	False	True
9a	2	2	0	False	True
9b	2	2	0	False	True
9c	2	2	0	False	True
11a	3	3	0	False	True
11b	3	3	0	False	True
11c	3	2	0	False	True
Erlotinib	1	2	0	False	True

^aBBB level, blood brain barrier level, 0 = very high, 1 = high, 2 = medium, 3 = low, 4 = very low.

^bSolubility level, 1 = very low, 2 = low, 3 = good, 4 = optimal.

^cAbsorption level, 0 = good, 1 = moderate, 2 = poor, 3 = very poor.

^dCYP2D6, cytochrome P2D6, TRUE = inhibitor, FALSE = non inhibitor.

^ePBB, plasma protein binding, FALSE means less than 90%, TRUE means more than 90%.

3. Conclusion

In this work new compounds having the essential pharmacophoric features of EGFR inhibitors were designed and synthesised. Three series of 1H-pyrrole, pyrrolo[3,2-d]pyrimidine and pyrrolo[3,2-e][1, 4]diazepine derivatives were obtained. Compound 9c, the fused pyrimido[4,5-b]pyrrolizine derivative, showed selectivity towards the HCT116 colon cancer cell line with activity comparable with that of DOX (IC₅₀ = 0.009 and 0.008 µM, respectively). While compound 8b, the 1H-pyrrole derivative featuring the extended 2-(piperidin-1-yl)acetamido moiety at C-6, showed a broad-spectrum inhibition against Hep3B, HCT116 and MCF-7cell lines (IC₅₀ = 0.049, 0.031 and 0.043 µM, respectively). Regarding compound 8b, the kinase profiling evaluations revealed its inhibitory activity against three types of kinases: CDK2/Cyclin A1, DYRK3 and GSK3 alpha. Therefore, this study reinforces the promising anticancer efficacy of 1H-pyrrole pyrrolo[3,2-d]pyrimidine derivatives based on their multi-targets activity, which were supported by molecular modelling characteristics. Add to that, their superior predicted safety and predicted pharmacokinetic properties as drug-like or lead-like molecules.

4. Experimental protocol

4.1. Chemistry

All details of chemical reagents and apparatus were described in Supplementary Data. Compounds 2⁶¹, 4a–c⁶³, 5a–c,⁶³ 7a–c and 10a–c⁷⁶ were prepared using the previously reported methods.

4.1.1. General method for preparation of compounds (6a–c)

To a solution of 4-nitrobenzaldehyde (3.75 mmol) in absolute ethanol (20 ml), the pyrrolizines 5a–c (3.75 mmol) and 0.5 ml of glacial acetic acid were added. The reaction mixture was refluxed for 4 h. The reaction mixture was left to cool. The products 6a–c were crystallised out as orange crystals and recrystallized from ethanol

4.1.2. (EZ)-7-Cyano-6-((4-nitrobenzylidene)amino)-N-phenyl)-2,3-dihydro-1H-pyrrolizine-5-carboxamide (6a)

The title compound was obtained from compound 5a as orange crystals, m.p. 278–80 °C, yield 76%, IRʋ_max/cm⁻¹3279 (NH), 3076 (C-H aromatic), 2935, 2852 (CH₂), 2210 (CN), 1658 (CO), 1601 (C = N), 1548, 1469 (C = C, NH), 1412, 1337 (C-N, C-O). ¹H NMR (CDCl₃, 400 MHz): δ 2.60 (m, 2H, CH₂-2), 3.07 (t, 2H, J = 7.4 Hz, CH₂-1), 4.57 (t, 2H, J = 7.4 Hz, CH₂-3), 7.17 (t, 1H, J = 8.2 Hz, CH-4′), 7.41 (t, 2H, J = 8.2 Hz, CH-3′, −5′), 7.62 (d, 2H, J = 8.2 Hz, CH-2′, −6′), 8.1 and 8.41 (two d, 4 H, J = 8 Hz, CH-2′', −3′', −4′', −5′', −6′'), 9.28 (s, H, N=CH) and 10.40 (s, 1H, NH, disappeared on deuteration). ¹³C NMR (CDCl₃, 100 MHz): δ 24.5, 25.4, 50.4, 76.7, 116.0, 119.1, 119.6, 124.4, 124.4, 129.1, 129.3, 137.8, 137.9, 140.9, 148.9, 149.7, 156.4, 158. MS (EI): m/z (%) 400 (M⁺+1, 13), 399 (M⁺, 35), 307 (100). Anal. Calcd. for C₂₂H₁₇N₅O₃ (399.40):C, 66.16; H, 4.29; N, 17.53. Found: C, 68.35; H, 4.15; N, 17.23.

4.1.3. (EZ)-7-Cyano-6-((4-nitrobenzylidene)amino)-N-(4-tolyl)-2,3-dihydro-1H-pyrrolizine-5-carboxamide (6b)

The title compound was obtained from compound 5b as orange crystals, m.p. 291–3 °C, yield 79%, IRʋ_max/cm⁻¹3280 (NH), 3069 (C-H aromatic), 2918 (CH₃, CH₂), 2210 (CN), 1657 (CO), 1601 (C = N), 1544, 1519 (C = C,N-H), 1411, 1338 (C-N, C-O). ¹H NMR (CDCl₃, 400 MHz): δ 2.37 (s, 3H, CH₃Ph), 2.60 (m, 2H, CH₂-2), 3.10 (t, 2H, J = 7.4 Hz, CH₂-1), 4.59 (t, 2H, J = 7.4 Hz, CH₂-3), 7.21 (d, 2H,J = 8.4 Hz, CH-3′, −5′), 7.53 (d, 2H,J = 8.4 Hz, CH-2′, −6′), 8.10 (d, 2H,J = 8.4 Hz, CH-2′', −6′'), 8.39 (d, 2H,J = 8.4 Hz, CH-3′', −5′'), 9.28 (s, H, N=CH) and 10.37 (s, 1H, NH, disappeared on deuteration). ¹³C NMR (CDCl₃, 100 MHz): δ 20.9, 24.5, 25.5, 50.3, 76.7, 116.1, 119.3, 119.6, 124.4, 129.1, 129.8, 134.1, 135.4, 137.7, 140.9, 148.7, 149.6, 156.3, 157.9.MS (EI): m/z (%) 414 (M ^{+ 1}, 15), 413 (M⁺, 51), 307 (100). Anal. Calcd. for C₂₃H₁₉N₅O₃ (413.43): C, 66.82; H, 4.63; N, 16.94. Found: C, 67.01; H, 4.58; N, 16.91.

4.1.4. (EZ)-N-(4-Chlorophenyl)-7-cyano-6-((4-nitrobenzylidene)amino)-2,3-dihydro-1H-pyrrolizine-5-carboxamide (6c)

The title compound was obtained from compound 5c as yellow crystals, m.p. 283–5 °C, yield 83%, IRʋ_max/cm⁻¹3245 (NH), 3111 (C-H aromatic), 2968 (CH₂), 2214 (CN), 1680 (CO), 1597 (C = N), 1541, 1482 (C = C, NH), 1412, 1344 (C-N, C-O), 836, 754 (C-Cl). ¹H NMR (CDCl₃, 400 MHz): δ 2.63 (m, 2H, CH₂-2), 3.11 (t, 2H, J = 7.4 Hz, CH₂-1), 4.58 (t, 2H, J = 7.4 Hz, CH₂-3), 7.38 (d, 2H,J = 8.4 Hz, CH-3′, −5′), 7.60 (d, 2H,J = 8.4 Hz, CH-2′, −6′), 8.10 (d, 2H,J = 8.4 Hz, CH-2′', −6′'), 8.42 (d, 2H, J = 8.4 Hz, CH-3′', −5′'), 9.31 (s, H, N=CH) and 10.45 (s, 1H, NH, disappeared on deuteration). ¹³C NMR (CDCl₃, 100 MHz): δ 24.5, 25.4, 50.4, 76.7, 115.9, 119.2, 119.8, 124.4, 129.1, 129.3, 135.4, 137.6, 137.7 140.9, 148.9, 149.7, 156.4, 157.9 MS (EI): m/z (%) 435 (M ^{+ 2}, 74), 433 (M⁺, 76), 405 (100). Anal. Calcd. for C₂₂H₁₆ClN₅O₃ (433.85):C, 60.91; H, 3.72; N, 16.14. Found: C, 61.10; H, 3.68; N, 16.11.

4.1.5. General method for preparation of compounds (8a–c)

A mixture of compounds 7a–c (2.91 mmol), piperidine (0.5 g, 5.9 mmol) and dry sodium bicarbonate (0.5 g, 5.9 mmol) in absolute ethanol (10 ml) was refluxed for 8 h. Then, the reaction mixture was filtered while hot, and the produced white crystals upon concentration were collected, dried, and recrystallized from ethanol.

4.1.5.1. 7-Cyano-N-phenyl-6–(2-(piperidin-1-yl)acetamido)-2,3-dihydro-1H-pyrrolizine-5-carboxamide (8a)

The title compound was obtained from compound 7a as white crystals, m.p. 205–7 °C, yield 72%, IRʋ_max/cm⁻¹3243, 3130 (NHs), 3060 (C-H aromatic), 2935, 2852 (CH₂), 2220 (CN), 1660 (COs), 1601, 1557, 1493 (C = C, NH), 1451, 1390, 1319 (C-N, C-O). ¹H NMR (CDCl₃, 400 MHz):δ 1.5 (m, 2H, CH₂-4′'), 1.7 (m, 4H, CH₂-3′', −5′'), 2.54 (m, 2H, CH₂-2), 2.69 (t, 4H,J = 5.2 Hz, CH₂-2′', −6′'), 3.02 (t, 2H, J = 7.6 Hz, CH₂-1), 3.27 (s, 2H, COCH₂), 4.40 (t, 2H, J = 7.2 Hz, CH₂-3), 7.12 (t, 1H, J = 8 Hz, CH-4′), 7.34 (t, 2H, J = 8 Hz, CH-3′, −5′), 7.6 (d, 2H, J = 8 Hz, CH-2′, CH-6′), 9.44 (s, H, NHCOCH₂, which disappeared on deuteration) and 9.80 (s, H, CONH phenyl, which disappeared on deuteration). ¹³C NMR (CDCl₃, 100 MHz): δ 23.5, 25, 25.7, 26.1, 49.6, 55.1, 62.1, 83.9, 114, 119.5, 120.3, 124.1, 124.4, 129, 138.3, 145.7, 157.6, 173.2. MS (EI): m/z (%) 392 (M⁺+1, 57), 391 (M⁺, 67), 138 (100).Anal. Calcd. forC₂₂H₂₅N₅O₂ (391.47): C, 67.50; H, 6.44; N, 17.89. Found: C, 67.80; H, 6.72; N, 17.67.

4.1.5.2. 7-Cyano-6–(2-(piperidin-1-yl)acetamido)-N-(4-tolyl)-2,3-dihydro-1H-pyrrolizine-5-carboxamide (8b)

The title compound was obtained from compound 7b as white crystals, m.p. 178–80 °C, yield 72%, IRʋ_max/cm⁻¹3240, 3186 (NHs), 3064 (C-H aromatic), 2983, 2930, 2853 (CH₃, CH₂), 2222 (CN), 1656 (COs), 1604, 1547, 1513 (C = C, NH), 1455, 1385, 1321 (C-N, C-O). ¹H NMR (CDCl₃, 400 MHz): δ 1.51 (m, 2H, CH₂-4′'), 1.72 (m, 4H, CH₂-3′' −5′'), 2.33 (s, 3H,CH₃Ph), 2.54 (m, 2H, CH₂-2), 2.70 (t, 4H,J = 5.2 Hz, CH₂-2′' −6′'), 3.02 (t, 2H, J = 7.4 Hz, CH₂-1), 3.28 (s, 2H, COCH₂), 4.39 (t, 2H, J = 7.2 Hz, CH₂-3),7.13 (d, 2H, J = 8 Hz, CH-3′, −5′), 7.49 (d, 2H, J = 8 Hz, CH-2′, −6′), 9.42 (s, H, NHCOCH₂, disappeared on deuteration) and 9.69 (s, H, CONH phenyl, disappeared on deuteration). ¹³C NMR (CDCl_3, 100 MHz): δ 20.9, 23.5, 25.1, 25.7, 26.1, 49.5, 55.1, 62, 83.9, 114.1, 119.5, 120.4, 124.2, 129.5, 133.8, 135.8, 145.6, 157.5, 173.2. MS (EI): m/z (%) 407 (M⁺+2, 69), 406 (M⁺+1, 82), 405 (M⁺, 68) 148 (100).Anal. Calcd. for C₂₃H₂₇N₅O₂(405.49): C, 68.13; H, 6.71; N, 17.27. Found: C, 68.33; H, 6.59; N, 17.50.

4.1.5.3. N-(4-Chlorophenyl)-7-cyano-6–(2-(piperidin-1-yl)acetamido)-2,3-dihydro-1H-pyrrolizine-5-carboxamide (8c)

The title compound was obtained from compound 7c as white crystals, m.p. 210–2 °C, yield 74%, IRʋ_max/cm⁻¹3233, 3179 (NHs), 3053 (C-H aromatic), 2986, 2932 (CH₂), 2222 (CN), 1658 (COs), 1602, 1546 (C = C, NH), 1489, 1456, 1387 (C-N, C-O), 829, 792 (C-Cl). ¹H NMR (CDCl₃, 400 MHz):δ 1.51 (m, 2H, CH₂-4′'), 1.74 (m, 4H, CH₂-3′' −5′'), 2.55 (m, 2H, CH₂-2), 2.74 (t, 4H, J = 5.2 Hz, CH₂-2′' −6′'), 3.02 (t, 2H, J = 7.4 Hz, CH₂-1), 3.32 (s, 2H, COCH₂), 4.39 (t, 2H, J = 7.4 Hz, CH₂-3), 7.28–7.58 (4H, aromatic protons), 9.41 (s, H, NHCOCH₂, disappeared on deuteration) and 9.93 (s, H, CONH phenyl, disappeared on deuteration). ¹³C NMR (CDCl₃, 100 MHz): δ23.4, 25, 25.6, 26.1, 49.5, 55, 62, 83.9, 113.9, 120, 120.7, 124.4, 129, 131.6, 136.9, 145.8, 157.6, 173.3. MS (EI): m/z (%) 426 (M⁺+1, 0.9), 98 (100). Anal. Calcd. for C₂₂H₂₄ClN₅O₂ (425.91): C, 62.04; H, 5.68; N, 16.44. Found: C, 62.30; H, 5.90; N, 16.15.

4.1.6. General method for preparation of compounds (9a–c)

A mixture of 6-amino-7-cyano-N-phenyl-2,3-dihydro-1H-pyrrolizine-5-carboxamide 5a–c (3.75 mmol) and excess triethyl orthoformate was refluxed for 12 h. The solvent was then removed using a rotary evaporator and the obtained residue was crystallised from ethanol.

4.1.6.1. Methyl-4-oxo-3-phenyl-4,6,7,8-tetrahydro-3H-pyrimido[4,5-b]pyrrolizine-9-carbonitrile (9a)

The title compound was obtained from compound 5a as white crystals, m.p. 243–5 °C, yield 55%, IRʋ_max/cm⁻¹3061 (C-H aromatic), 2966, 2918 (CH₃, CH₂), 2214 (CN), 1693 (CO), 1524 (C = C), 1421, 1304 (C-N, C-O).¹H NMR (CDCl₃, 400 MHz): δ 2.26 (s, 3H, CH₃C = N), 2.67 (m, 2H, CH₂-7), 3.19 (t, 2H, J = 7.4 Hz, CH₂-8), 4.40 (t, 2H, J = 7.2 Hz, CH₂-6),7.25 (d, 2H, J = 8 Hz, CH-2′, −6′) and 7.51–7.59 (m, 3H, CH-3′, −4′, −5′). ¹³C NMR (CDCl₃, 100 MHz): δ 24.3, 25.1, 26.2, 48, 81.1, 113.4, 113.9, 128, 129.4, 130, 137.3, 148.4, 152.3, 154.1, 154.3. MS (EI): m/z (%) 291 (M⁺+1, 3), 290 (M⁺, 15), 85 (100). Anal. Calcd. for C₁₇H₁₄N₄O (290.32): C, 70.33; H, 4.86; N, 19.30. Found: C, 70.12; H, 4.90; N, 19.41.

4.1.6.2. 2-Methyl-4-oxo-3–(4-tolyl)-4,6,7,8-tetrahydro-3H-pyrimido[4,5-b]pyrrolizine-9-carbonitrile (9b)

The title compound was obtained from compound 5b as white crystals, m.p. 251–3 °C, yield 59%, IRʋ_max/cm⁻¹3073 (C-H aromatic), 2924 (CH₃, CH₂), 2208 (CN), 1694 (CO), 1591 (C = C), 1427, 1302 (C-N, C-O). ¹H NMR (CDCl₃, 400 MHz): δ 2.27 (s, 3H, CH₃C = N), 2.46 (s, 3H, CH₃Ph), 2.66 (m, 2H, CH₂-7), 3.18 (t, 2H, J = 7.6 Hz, CH₂-8), 4.41 (t, 2H, J = 7.2 Hz, CH₂-6), 7.12 and 7.37 (two d, 4 H, J = 8 Hz, p-substituted phenyl ring). ¹³C NMR (CDCl₃, 100 MHz): δ21, 24.3, 25.1, 26.2, 48, 81.1, 113.5, 114, 127.9, 130.5, 137.2, 139.4, 148.5, 152.3, 154.1, 154.3. MS (EI): m/z (%) 305 (M⁺+1, 7), 304 (M⁺, 51), 122 (100). Anal. Calcd. for C₁₈H₁₆N₄O (304.35): C, 71.04; H, 5.30; N, 18.41. Found: C, 70.78; H, 5.50;N, 18.30.

4.1.6.3. 3–(4-Chlorophenyl)-2-methyl-4-oxo-4,6,7,8-tetrahydro-3H-pyrimido[4,5-b]pyrrolizine-9-carbonitrile (9c)

The title compound was obtained from compound 5c as white crystals, m.p. 268–70 °C, yield 60%, IRʋ_max/cm⁻¹3060 (C-H aromatic), 2924 (CH₃, CH₂), 2223 (CN), 1689 (CO), 1533 (C = C), 1419, 1300 (C-N, C-O), 840, 774 (C-Cl). ¹H NMR (CDCl₃, 400 MHz): δ 2.25 (s, 3H, CH₃C = N), 2.67 (m, 2H, CH₂-7), 3.19 (t, 2H, J = 7.2 Hz, CH₂-8), 4.39 (t, 2H, J = 7 Hz, CH₂-6), 7.19 and 7.54 (two d, 4 H, J = 8 Hz, p-substituted phenyl ring). ¹³C NMR (CDCl_3, 100 MHz): δ 24.4, 25.2, 26.2, 48.1, 81.3, 113.3, 113.8, 129.5, 130.3, 135.6, 135.8, 148.4, 152.6, 153.7, 154.2. MS (EI): m/z (%) 324 (M⁺, 1), 91 (100).Anal. Calcd. for C₁₇H₁₃ClN₄O (324.76): C, 62.87; H, 4.03; N, 17.25. Found: C, 62.69; H, 3.97; N, 17.55.

4.1.7. General method for preparation of compounds (11a–c)

To a solution of diazepino[5,6-b]pyrrolizine-10-carbonitrile derivative 10a–c (3.26 mmol) in acetone, ethyl chloroacetate (0.4 g, 3.26 mmol) and anhydrous potassium carbonate (0.45 g, 3.26 mmol) was added. After refluxing for 6 h, the mixture was filtered and left to cool. The obtained white crystals were collected and recrystallized from ethanol.

4.1.7.1. Ethyl-2–(10-cyano-2,5-dioxo-4-phenyl-2,3,4,5,8,9-hexahydro-[1, 4]diazepino[5,6-b]pyrrolizin-1(7H)-yl)acetate (11a)

The title compound was obtained from compound 10a as white crystals, m.p. 179–81 °C, yield 75%, IRʋ_max/cm⁻¹3059 (C-H aromatic), 2977 (CH₂), 2220 (CN), 1740, 1693, 1643 (COs), 1549 (C = C), 1461, 1369, 1306 (C-N, C-O). ¹H NMR (CDCl₃, 400 MHz): δ 1.29 (t, 3H, J = 7.2 Hz, CH₃CH₂),2.59 (m, 2H, CH₂-8), 3.08 (t, 2H, J = 7.4 Hz, CH₂-9), 4.24 (q, 2H, J = 7.2 Hz, CH₂CH₃), 4.39 (t, 2H, J = 7.2 Hz, CH₂-7), 4.49 (s, 2H, CH₂-3), 4.76 (s, 2H, NCH₂CO) and 7.29–7.46 (m, 5H, aromatic protons). ¹³C NMR (CDCl₃, 100 MHz): δ 14, 25.1, 25.4, 48.3, 49.3, 54.5, 61.9, 79.1, 114.2, 114.5, 125.7, 127.2, 129.2, 134.2, 141.6, 148.4, 159.6, 166.5, 167.5. MS (EI): m/z (%) 393 (M ^{+ 1}, 1), 392 (M⁺, 1), 106 (100).Anal. Calcd. for C₂₁H₂₀N₄O₄(392.41): C, 64.28; H, 5.14; N, 14.28. Found: C, 64.58; H, 5.36; N, 14.25.

4.1.7.2. Ethyl-2–(10-cyano-2,5-dioxo-4–(4-tolyl)-2,3,4,5,8,9-hexahydro-[1, 4]diazepino[5,6-b]pyrrolizin-1(7H)-yl)acetate (11b)

The title compound was obtained from compound 10b as white crystals, m.p. 213–5 °C, yield 71%, IRʋ_max/cm⁻¹3077 (C-H aromatic), 2970, 2924 (CH₃, CH₂), 2220 (CN), 1741, 1690, 1642 (COs), 1548, 1465 (C = C), 1371, 1307 (C-N, C-O). ¹H NMR (CDCl₃, 400 MHz): δ 1.28 (t, 3H, J = 7.2 Hz, CH₃CH₂), 2.37 (s, 3H, CH₃Ph),2.58 (m, 2H, CH₂-8), 3.07 (t, 2H, J = 7.4 Hz, CH₂-9), 4.24 (q, 2H, J = 7.2 Hz, CH₂CH₃), 4.40 (t, 2H, J = 7.2 Hz, CH₂-7), 4.46 (s, 2H, CH₂-3), 4.75 (s, 2H, NCH₂CO), 7.21 and 7.30 (two d, 4 H, J = 8.4 Hz, p-substituted phenyl ring). ¹³C NMR (CDCl₃, 100 MHz): δ 14, 21, 25.1, 25.4, 48.3, 49.3, 54.6, 61.9, 79.1, 114.2, 114.6, 125.5, 129.8, 134.2, 137.2, 139, 148.2, 159.7, 166.5, 167.6.MS (EI): m/z (%) 407 (M⁺+1, 14), 406 (M⁺, 52), 186 (100). Anal. Calcd. for C₂₂H₂₂N₄O₄(406.43): C, 65.01; H, 5.46; N, 13.78. Found: C, 65.30; H, 5.71; N, 13.77.

4.1.7.3. Ethyl-2–(4-(4-Chlorophenyl)-10-cyano-2,5-dioxo-2,3,4,5,8,9-hexahydro-[1, 4]diazepino[5,6-b]pyrrolizin-1(7H)-yl)acetate (11c)

The title compound was obtained from compound 10c as white crystals, m.p. 220–1 °C, yield 78%, IRʋ_max/cm⁻¹3092 (C-H aromatic), 2970 (CH₂), 2222 (CN), 1749, 1686, 1647 (COs), 1549 (C = C), 1490, 1471, 1373, 1306 (C-N, C-O), 838, 772 (C-Cl). ¹H NMR (CDCl₃, 400 MHz): δ 1.28 (t, 3H, J = 7.2 Hz, CH₃CH₂), 2.60 (m, 2H, CH₂-8), 3.09 (t, 2H, J = 7.6 Hz, CH₂-9), 4.24 (q, 2H, J = 7.2 Hz, CH₂CH₃), 4.39 (t, 2H, J = 7.2 Hz, CH₂-7), 4.49 (s, 2H, CH₂-3), 4.75 (s, 2H, NCH₂CO), 7.38 and 7.43 (two d, 4 H, J = 8.4 Hz, p-substituted phenyl ring). ¹³C NMR (CDCl₃, 100 MHz): δ 14.1, 25.2, 25.5, 48.4, 49.4, 54.4, 62, 79.3, 114.1, 114.4, 126.9, 129.4, 132.8, 134.4, 140, 148.6, 159.6, 166.4, 167.5. MS (EI): m/z (%) 428 (M⁺+2, 23), 426 (M⁺, 64), 186 (100). Anal. Calcd. for C₂₁H₁₉ClN₄O₄(426.85): C, 59.09; H, 4.49; N, 13.13. Found: C, 59.32; H, 4.39; N, 13.10.

4.2. Biological evaluation

4.2.1. In vitro anticancer screening

Cytotoxic activities evaluation of the synthesised compounds were done using the sulforhodamine B (SRB) method following the previous reported method⁶⁵ described in Supplementary data.

4.2.2. Kinase profiling assay

Compound 8b was selected to evaluate its inhibitory activity against 20 kinases according to the reported method.⁶⁶ The kinases inhibition assay was done by KINEXUS Corporation, Vancouver, BC, Canada, using the radiolabeled ATP determination method. Imatinib was used as a reference drug and blank control was set up and the corrected activity for the protein kinase target was determined. The results were presented as % inhibition, Table 3.

4.2.3. Perturbation of cell cycle analysis

Cell cycle distribution analysis was performed for compounds 6a and 8b based on the previously described method in Supplementary data⁷⁷.

4.3. In silico studies

4.3.1. Docking studies

Crystallographic structures of EGFR CDK-2, and DYRK3, and GSK3 were retrieved from Protein Data Bank [PDB ID: 4HJO, resolution 2.75 Å and PDB ID: 6GUH, resolution 1.50 Å, PDB: 5Y86, resolution 1.90 Å, PDB: 5HLP, resolution 2.45 Å respectively] (http://www.pdb.org), and considered as targets for docking simulations. The docking analysis was performed using MOE software^78–80 to evaluate the free energies and binding mode of the designed molecules against EGFR^, CDK-2 and DYRK3, and GSK3. At first, the crystal structures of EGFR, CDK-2 and DYRK3, and GSK3 were prepared by removing water molecules and retaining only one chain and their co-crystallised ligands, erlotinib and AZD5438, HRM, and 65 A, respectively. Then, the protein structures were protonated, and the hydrogen atoms were hidden. Next, the energy was minimised, and the binding pockets of the protein were defined.

The 2D structures of the synthesised compounds and the co-crystallised ligands, erlotinib and AZD5438, HRM, and 65 A were sketched using ChemBioDraw Ultra 14.0 and saved as MDL-SD format. Then, the saved files were opened using MOE and 3 D structures were protonated. Next, energy minimisation was applied. Before docking the synthesised compounds, validation of the docking protocol was carried out by running the simulation only using the co-crystallised ligands and low RMSD between docked and crystal conformations. The molecular docking of the synthesised compounds and the co-crystallised ligands was performed using a default protocol. MOE docking parameters include Triangle Matcher Algorithm with two rescoring functions London dG and GBVI/WSA dG were utilised to generate 10 poses of each compound. As a result, mdb output files were generated enclosing all docking results with scoring and multiple conformations of ligands. Results were finally inspected to determine the most promising compound by visualising various interactions of ligands within the binding pocket. The output from MOE was further analysed and visualised using Discovery Studio 4.0 software. The output from MOE was further analysed and visualised using Discovery Studio 4.0 software^47,48,56,60. Mutations, missing regions and active/inactive states of the receptors were presented in Supplementary data.

4.3.2. In silico toxicity prediction

The toxicity parameters of the synthesised compounds were calculated using Discovery studio 4.0 as described in Supplementary data^1,57.

4.3.3. In silico ADMET studies

ADMET descriptors (absorption, distribution, metabolism, excretion, and toxicity) of the synthesised compounds were determined using Discovery studio 4.0 as described in Supplementary data.

Disclosure statement

All authors of the above manuscript have not declared any conflict of interest.

Additional information

Funding

Deanship of scientific research at Umm Al-Qura University supported this work by grant code (22UQU4290565DSR37).