Synthesis and in vitro anticancer activity of some 2-oxindoline derivatives as potential CDK2 inhibitors

Abstract Novel series of 2-oxindoline hydrazones 6a–h, 3-hydroxy-2-oxoindolines 9a–d and 2-oxoindolin-3-ylidenes 10a–d were prepared and assessed for their anticancer activity towards breast cancer cell line (MCF7). Compounds 6c, 6d, 6g, 9d, 10a and 10b (IC50 = 14.0 ± 0.7, 15.6 ± 0.7, 13.8 ± 0.7, 4.9 ± 0.2, 6.0 ± 0.3 and 10.8 ± 0.5 µM, respectively) showed the highest growth inhibition activity against MCF7 when compared to staurosporine (IC50 = 14.5 ± 0.7 µM). Cell cycle analysis exposed arrest at G1 phase for compounds 6c, 10 and 10b, at S phase for compounds 6d and 9d, and at G1/S phase for compound 6g. Apoptotic effect of compounds 6c, 6d, 6g, 9d, 10a and 10b was confirmed via their early and late apoptotic effects. A safety profile was revealed for compounds 6c, 6d, 6g, 9d, 10a and 10b on MCF10A treated normal cell. Also, compounds 6c and 10b displayed a promising CDK2 inhibition activity (IC50 = 0.22 ± 0.01, 0.25 ± 0.01 µM, respectively). Also, docking study revealed comparable interactions with the native ligand (5-bromoindirubin). ADMET computational studies forecast the promising pharmacokinetic profile of the targeted compounds. Communicated by Ramaswamy H. Sarma


Introduction
Cancer is represented as one of the major causes of death in the globe (Cancer, WHO, 2020).Despite the progresses made in the cancer therapy, various limitations still occur including drug resistance (Wu et al., 2014), several side effects (Decalf et al., 2017) and the selectivity issue towards the cancer cells rather the normal cells (Shafei et al., 2017).Thus, designing new compounds that are active and selective still represents a main challenge (Iyer, 2015;Lin et al., 2019).Higher rates of proliferation and evading apoptosis that is closely related to cell cycle are listed as common cancer hallmarks (Zhong et al., 2020).Cyclin-Dependent Kinases (CDKs) have gotten a great attention as key cell cycle regulators (Kargbo, 2022;Laphanuwat & Jirawatnotai, 2019;Martin et al., 2017;Peyressatre et al., 2015;Said et al., 2022).Association of CDKs with their cyclins causes their activation which catalyzes the phosphate function transfer from adenosine triphosphate (ATP) to amino acids threonine or serine in the protein bases and any abnormality of CDKs is connected to the pathogenesis of malignancies (Martin et al. 2017;Asghar et al. 2015).CDK2 is one of the key controllers for the cell cycle and it directs the G1/S and G2/M phases in cell cycle (Chung & Bunz, 2010;De Boer et al., 2008;Donjerkovic & Scott, 2000;Hu et al., 2001), and takes a part in the apoptosis cascade (Gil-G� omez, 2004).Accordingly, inhibition the activity of CDK2 could produce apoptosis and cell cycle detention at the latter phases, providing an acceptable strategy for cancer therapy (Guardavaccaro & Pagano, 2006).
Guided by the previous facts, our rational was directed for synthesis of some new oxindoles as prospective antitumor agents against MCF7 breast cancer cells with expected CDK2 inhibition activity.The abovementioned binding mode of 5bromoindirubin (Figure 1) has provided valuable information on the essential interactions between that kinase type and the indirubin analogue (Ib).The latter mode of interaction also stimulated our interest to synthesis the title novel oxindoles 6a-h, 9a-d and 10a-d (Figure 2) with an expected near interaction pattern.Keeping in consideration maintaining the main core essential features (NH & O of the oxindole core), together with extending that core at position 3 with variable side chains aiming to furnish extra binding with the active site.Additionally, substitution at position 5 of the targeted oxindoline core was considered for some electronic and lipophilic diversity (Figure 2).
The designed compounds will be assessed for their anticancer activity against MCF7.The highest active compounds will be explored for CDK2 enzyme inhibition activity, cell cycle analysis, apoptosis assay, and also, safety profile valuation on normal cell line.Molecular docking study for CDK2 is aimed to predict the binding profiles of the investigated uppermost active compounds.Finally, ADMET computational study will be considered to predict their pharmacokinetics and toxicity of 6a-h, 9a-d and 10a-d.
The key intermediates 4a, b were refluxed for 3 h with the appropriate indolines 5a-d in absolute EtOH and AcOH to afford 2-oxoindolin-3-ylidene hydrazones 6a-h.The IR of 6a-h appeared the absorption bands of amide, indoline and hydrazide NHs in the region 3550-3100 cm À 1 whereas the bands of three carbonyl groups appeared around 1731-1612 cm À 1 .The 1 HNMR of 6a-d revealed the signal of -CH 2 around d 4.10-4.54ppm whereas 6e-h revealed the doublet signals due to the methyl group protons (-CH-CH 3 ) around d 1.48 ppm and the signal of chiral carbon proton (-CH-CH 3 ) around d 4.66-5.43ppm.In addition, the 1 HNMR of 6a-h revealed the D 2 O exchangeable signals of amide, indoline and hydrazide NHs at d 8. 29-9.21, 10.81-11.34 and 11.21-13.49ppm, respectively. 13CNMR of 6a-d exhibited the signals of -CH 2 around 55.00 ppm whereas those of 6e-h exhibited the signals of alanine -CH 3 and -CH-around d 16.64-16.99and 56.83-64.69ppm, respectively.Furthermore, 13 CNMR of 6a-h showed the signals of three signals of amide, indoline and hydrazide carbonyl carbons around d 162.73-165.13, 164.76-167.03 and 167.31-174.35ppm, respectively.The mass spectra of compounds 6a-h appeared the ion peak of their molecular weight.On the other side, N-(4-acetylphenyl)benzamide ( 4), which prepared by the reaction on benzoyl chloride with 4-aminoacetophenone (7) in DCM in the presence of TEA (Yadlapalli et al., 2013

Structure activity relationship (SAR study).
As presented in Table 1, relating the observed IC 50 values of the 2oxindoline hydrazones 6a-h to their structural features revealed the impact of introduced halogen substituents at position 5 on their activities (Figure 3).Moreover, substitution with -CH 3 group at the side chain of compounds 6a-h seems less promising than the un-substitution (Figure 3).Concerning the 3-hydroxy-2-oxoindolines 9a-d, the order of activity regarding the halogen substitution at position 5 is decreased as follow 'Br > F > Cl'.Meanwhile, for 2-oxoindolin-3-ylidenes 10a-d the order of activity decreased as follow 'F > Cl > Br'.Also, comparing the activity of series 9a-d and   10a-d, the arylidene compounds 10a-c overtop the activity of the flexible analogs 9a-c except 10d (Figure 3).

Cell cycle and apoptosis 2.2.2.1. Cell cycle.
Derivatives 6c, 6d, 6 g, 9d, 10a and 10b showing the topmost antiproliferative activity against MCF7 were picked for more mechanistic examination.The effects of the tested compounds on the progression of cell cycle and induction of apoptosis were assessed.Flow cytometry an analysis utilizing BD FASCC alibur was performed (Tolosa et al., 2015).The higher cell accumulations (64.62, 59.52 and 63.90%, respectively) were noticed at the G0-G1 phase for compounds 6c, 10a and 10b (control; 53.82%) signifying cell cycle arrest at G1 phase (Table 2 and Figure 4).Moreover, cells accumulations were also observed at S phase for compounds 6d and 9d (52.11 and 43.58 vs. control; 34.69%) indicating the arresting of cell cycle at the S phase.Additionally, the treatment of compound 6g with MCF7 cells resulted in cells accumulation at G0-G1 and S phases (55.44 and 39.01%, respectively) when compared with the control cells (53.82% and 34.69%, respectively) representing the arresting of cell cycle at the G1/S phase.Compounds 6c, 6d, 6 g, 9d, 10a and 10b inhibit the proliferation of MCF7 cancer cell, and cause apoptotic DNA fragmentation.

Apoptosis assay.
Annexin V/PI analysis of compounds 6c, 6d, 6g, 9d, 10a and 10b to induce apoptosis in early and late apoptosis showed a significant level of MCF7 apoptosis in comparison with the control cells (Table 3 and Figure 5).

Cytotoxicity of tested compounds on normal cell line
Compounds 6c, 6d, 6g, 9d, 10a and 10b the most inhibitors for MCF7 growth, were evaluated for their cytotoxicity against MCF10A human breast epithelial normal cell line comparing with staurosporine as reference compound using MTT assay.The safety profile of the tested compounds is expressed in cytotoxic concentration (CC 50 ).In Table 4, the toxicity of compounds 6c, 6d, 6g, 9d, 10a and 10b on MCF10A compared to staurosporine are illustrated.The selectivity index (SI) of 6c, 6d, 6g, 9d, 10a and 10b was calculated by their selectivity towards MCF7 infected cancer cell rather MCF10A (selectivity index ¼ CC 50 on MCF10A/IC 50 on MCF7; El-Miligy et al., 2018).Compounds 6c, 6d, 6g, 9d, 10a and 10b revealed selectivity index higher than that of staurosporine towards MCF7 (Table 4).

Molecular docking
The noticed CDK2 inhibitory activities attained by the targeted oxindoles 6c, 6d, 6g, 9d, 10a and 10b encouraged us for the exploration of their binding pattern within the binding pocket of CDK2.The crystal structure '2BHE' has the 5bromoindirubin Ib, an oxindole analogue with promising CDK2 inhibition activity, co-crystalized with CDK2, consequently, it has been chosen for our docking study.At first, the validation step results in RMSD ¼ 0.701 Å, docking score ¼ À 11.479 Kcal/mol, and a similar binding profile to that of the native ligand (Figure 6).The CDK2 inhibitory activity of compound (Ib) was attributed to its binding in the pocket taken by the ATP of adenine ring.Three H interactions were displayed with the CDK2 skeleton at the hinge zone.The amidic NH and O of compound (Ib) create Hbonds with NH group of Leu-83 and oxygen of Glu-81 whereas a 3rd H-bond is created between NH of (Ib) and the Leu-83 oxygen (Figure 6).The results of docking for the targeted compounds 6c, 6d, 6g, 9d, 10a and 10b showed a similar binding mode with docking scores are comparable to that of 5-bromoindirubin (Ib) (Supporting data).
A closer look on the binding modes of compounds 6c and 10b, the topmost CDK2 inhibitors; IC 50 ¼ 0.22 ± 0.01, 0.25 ± 0.01 mM, respectively) revealed the next.Figure 7 showed the binding mode of 6c 'docking score (S) ¼ À 11.266 kcal/mol' involves a network of H interactions with the essential residues in the binding site.Two H-bonds were mediated with Leu83 residue via the C ¼ O group of the oxindole core and the side chain CH 2 group.In addition, one hydrogen bond was displayed with Glu81 via the NH group of the oxindole core, and another one with Phe82 through the C ¼ O group of the oxindole core.Furthermore, an arene-H hydrophobic interaction with Gln85 residue was displayed through the side chain phenyl moiety.
Also, compound 10b mediates two H-bonds with Leu83 and Phe82 residues through the side chain and oxindole C ¼ O, respectively (Figure 8).Moreover, arene-H hydrophobic interactions are appeared with Val18 and Phe80 through the oxindole core.Besides, compound 10b achieved a low docking score (S) value of À 11.463 kcal/mol.
Its worthy mentioned that the lower docking score means better affinity for compound 6c and 10b, the topmost CDK2 inhibitors, with the binding site of the enzyme.Also, their docking scores are agreed with their IC 50 values which gave an expected explanation for their anticancer activity.

In silico ADMET study 2.3.2.1. In silico ADME predictions.
Swiss ADME is used for prediction of the pharmacokinetics and drug-likeness characteristics of the tested compounds.Blood-brain barrier (BBB), human gastrointestinal absorption (HIA), penetration, glycoprotein (P-gp) permeability non-substrate or substrate and, interactions with cytochromes P450 isomers (CYP) were expected.The bioavailability scores were calculated depending on 5 rules (Daina et al., 2017), Lipinski (Lipinski et al., 2001), Veber (Veber et al., 2002), Ghose (Ghose et al., 1999), Egan (Egan et al., 2000) and Muegge (Muegge et al., 2001) rules.Figure 9 showed the results as BOILIED-EGG, which is a 2D plot presented via the calculated TPSA and LogP properties of the checked compounds.The white region specifies the GIT passive absorption possibility; compounds 6a-h and 9a-d are located in the white zone.The yellow region forecasts the BBB permeation probability; compound 10a-d is located in the yellow part.Also, compounds 6a-h and 9a-d might not be substrates for the P-glycoprotein (PGPÀ ) (red colored dots) and thus reducing the opportunity of its efflux by the cancer cells (Campos et al., 2013), while compounds 10a-d expected to be (PGPþ) (blue colored dots) that might be effluxed via P-glycoprotein.Additionally, most of compounds predicted to display non-inhibitory activities on cytochrome P450 different isomers and thus are expected to mediate no drug-drug interactions when administered (Lynch & Price, 2007).Bioavailability scores were also predicted depending on the agreement of the compounds to Lipinski, Veber, Ghose, Egan and Muegge rules.A good oral bioavailability scores (0.55) were predicted for compounds 6a-h, 9a-d and 10a-d, whereby all the checked compounds revealed no violations from the abovementioned rules (Table 6).

Toxicity prediction.
Osiris Property was used to estimate the probable toxicities of the final compounds 6a-h, 9a-d and 10a-d.Prediction is based on the functional group semblance of the checked compound with the wide in vivo and in vitro studied compounds comprised in its database.Probability of low toxicity is coded by green color, mild toxicity is coded by yellow color, and high toxicity is coded by red color (Sander, 2001;Sander et al., 2009).The results showed that 6a-h, 9a-d and 10a-d are predicted to be safe with no toxicity related to tumorigenicity, mutagenicity, irritation and reproduction.

Conclusion
Three series of indoline hydrazones 6a-h, 3-hydroxy-2-oxoindolines 9a-d and 2-oxoindolin-3-ylidenes 10a-d were synthesized, and assessed for their antiproliferative activity towards MCF7 cancer cell.Compounds 6c, 6d, 6g, 9d, 10a and 10b exerted the highest inhibition activity against MCF7 when compared with staurosporine.Cell cycle analysis revealed MCF7 cell cycle arrest at G1 phase for compounds 6c, 10 and 10b, at S phase for compounds 6d and 9d and at G1/S phase for compound 6 g.Moreover, compounds 6c, 6d, 6g, 9d, 10a and 10b are apoptotic inducers within MCF7 treated cells.Good safety profile for 6c, 6d, 6g, 9d, 10a and 10b on MCF10A normal cell line was displayed.Additionally, compounds 6c and 10b displayed the highest CDK2 inhibition activity.Molecular docking study revealed a similar binding behavior of the tested compounds to that of the native ligand .Finally, ADMET studies predict the targeted compounds promising pharmacokinetic profile.

Figure 4 .
Figure 4. Cell cycle analysis for the examined compounds.
3092 cm À 1 whereas the bands of 3C ¼ O appeared at 1725-1652 cm À 1 .The 1 HNMR of 10a-d exhibited the signals of olefinic protons (-CH¼C-) at d 7.82 ppm in addition to the D 2 O exchangeable amide and indoline NHs signal at d 10.67 and 10.94 ppm, respectively.

Figure 6 .
Figure 6.2D (A) and 3D (B) ligand interaction diagram of the native ligand Ib in green and its redocked pose (red) in the binding site of CDK2-ATP.

Figure 9 .
Figure 9. HIA and BBB plot for the studied compounds.

Table 1 .
Anticancer activity (IC 50 mM) of the examined compounds and staurosporine towards MCF7 cancer cell.

Table 2 .
Cell cycle analysis for the examined compounds.

Table 4 .
CC 50 a (mM) and SI b of 6c
a CC 50 : The concentration reduces the cell viability by 50%.b SI: Selectivity index ¼ CC 50 on MCF10A/IC 50 on MCF7.

Table 6 .
ADME study of the tested compounds.