A Series of Isatin-Hydrazones with Cytotoxic Activity and CDK2 Kinase Inhibitory Activity: A Potential Type II ATP Competitive Inhibitor

Isatin derivatives potentially act on various biological targets. In this article, a series of novel isatin-hydrazones were synthesized in excellent yields. Their cytotoxicity was tested against human breast adenocarcinoma (MCF7) and human ovary adenocarcinoma (A2780) cell lines using MTT assay. Compounds 4j (IC50 = 1.51 ± 0.09 µM) and 4k (IC50 = 3.56 ± 0.31) showed excellent activity against MCF7, whereas compound 4e showed considerable cytotoxicity against both tested cell lines, MCF7 (IC50 = 5.46 ± 0.71 µM) and A2780 (IC50 = 18.96± 2.52 µM), respectively. Structure-activity relationships (SARs) revealed that, halogen substituents at 2,6-position of the C-ring of isatin-hydrazones are the most potent derivatives. In-silico absorption, distribution, metabolism and excretion (ADME) results demonstrated recommended drug likeness properties. Compounds 4j (IC50 = 0.245 µM) and 4k (IC50 = 0.300 µM) exhibited good inhibitory activity against the cell cycle regulator CDK2 protein kinase compared to imatinib (IC50 = 0.131 µM). A molecular docking study of 4j and 4k confirmed both compounds as type II ATP competitive inhibitors that made interactions with ATP binding pocket residues, as well as lacking interactions with active state DFG motif residues.


Introduction
Development of anticancer drugs is essential due to the increasing number of morbidity and mortality by cancer day-by-day all over the world. According to the International Agency for Research on Cancer, in 2018, around 18 million people were infected; 9.6 million people among them had died due to life threatening cancer [1,2]. It is rather alarming that cancer morbidity cases may increase to 29.5 million by 2040 [3]. Having said that, it is very much challenging to develop an anticancer drug due to the long and expensive synthesis/isolation process and the huge lack of opportunities to conduct clinical trials. Moreover, most of the anticancer drugs currently available are lacking specificity and have adverse effects. In this context, developing novel anticancer agents with great efficacy and

Cytotoxicity
The cytotoxicity of the synthesized compounds 4a-k was evaluated against two different cancer cell lines, namely MCF7 and A2780, and the results are summarized in Table 1. Among the tested compounds, the isatin-hydrazone 4j exhibited the highest inhibitory activity against MCF7 cell lines (1.51 ± 0.09 µM). It should be noted that 4k (3.56 ± 0.31), 4e (5.46 ± 0.71), 4i (7.77 ± 0.008) and 4f (9.07 ± 0.59) showed moderate inhibitory activity against MCF7 cell lines. In the case of A2780 cell lines, however, only the halogen-substituted compounds 4e, 4j, 4k and 4f showed a little inhibitory activity. Nevertheless, all of the tested compounds were more sensitive towards MCF7 compared to A2780 cell lines.  Figure 2 shows the dose-response curves for compounds 4j and 4k, which were the most cytotoxic compounds against the breast cancer cells lines (MCF7) at a concentration of 1.51 and 3.56 µM, respectively. The IC 50 values interpolated from dose-response data with five different concentrations were 0.1, 1, 10, 25 and 50 µM.
The cytotoxicity of the synthesized compounds 4a-k was evaluated against two different cancer cell lines, namely MCF7 and A2780, and the results are summarized in Table 1. Among the tested compounds, the isatin-hydrazone 4j exhibited the highest inhibitory activity against MCF7 cell lines (1.51 ± 0.09 μM). It should be noted that 4k (3.56 ± 0.31), 4e (5.46 ± 0.71), 4i (7.77 ± 0.008) and 4f (9.07 ± 0.59) showed moderate inhibitory activity against MCF7 cell lines. In the case of A2780 cell lines, however, only the halogen-substituted compounds 4e, 4j, 4k and 4f showed a little inhibitory activity. Nevertheless, all of the tested compounds were more sensitive towards MCF7 compared to A2780 cell lines.  Figure 2 shows the dose-response curves for compounds 4j and 4k, which were the most cytotoxic compounds against the breast cancer cells lines (MCF7) at a concentration of 1.51 and 3.56 μM, respectively. The IC50 values interpolated from dose-response data with five different concentrations were 0.1, 1, 10, 25 and 50 μM.

Structure-Activity Relationships (SARs) Study of 4a-k
The SARs study revealed that the cytotoxicity of 4a-k increased or decreased in the same fashion as increases or decreases in halogen substitution in the aromatic C-ring. It is also related to the position of the substituents. As depicted in Figure 3, the bromo substituent at 4-position of the C-ring gave IC50 = 15.7 μM against MCF7 cell lines (Table 1,

Structure-Activity Relationships (SARs) Study of 4a-k
The SARs study revealed that the cytotoxicity of 4a-k increased or decreased in the same fashion as increases or decreases in halogen substitution in the aromatic C-ring. It is also related to the position of the substituents. As depicted in Figure 3, the bromo substituent at 4-position of the C-ring gave IC 50 = 15.7 µM against MCF7 cell lines (Table 1, entry 4g), while at 3-position, it increased to IC 50 = 9.07 µM (Table 1, entry 4f). Interestingly, the bromo substituent's cytotoxicity at 2-position, increased dramatically to IC 50 = 5.46 µM (Table 1, entry 4e). Surprisingly, while 2-and 6-positions of the C-ring having respective chloro-and fluoro-substituents, the IC 50 of compound 4k was 3.56 µM (Table 1, entry 4k). More surprisingly, with both 2-and 6-positions of the C-ring with chloro-substituents, compound 4j exhibited the highest cytotoxicity of IC 50 entry 4a). On the other hand, A2780 cell lines were inhibited by the halogenated derivatives 4e, 4j, 4k and 4f. In this case, 2-bromo substituted derivatives showed higher activity than the other three.

CDK2 Protein Kinase Inhibitory Activity of 4a-k
The promising cytotoxicity of 4, especially 4j and 4k, motivated us to study further inhibitory activities against CDK2 protein kinase. As summarized in Table 2, 4j and 4k exhibited good inhibitory activity against cyclin-dependent kinase 2 (CDK2), which is half of that of the known kinase inhibitor imatinib.

In Silico Drug Likeness Property Analysis
Rational drug designing is the most significant part in modern drug discovery approaches. In this regard, computational ADME (absorption, distribution, metabolism and excretion) analysis can help us select the best drug in terms of cost, time and efficiency. Applying computational chemistry tools, in vitro and in vivo ADME prediction is now much more convenient and it can aid pharmaceutical industries to screen thousands of compounds within a short time [53]. Here, synthesized compounds (4a-k) were screened for predicted ADME values and the results are summarized in Table 3. Since high molecular weight compounds are always less effective in terms of intestinal absorption [54,55], our designed and synthesized isatin-hydrazones' (4a-k) molecular weights were kept low, in between 263-328 Da. Compounds 4a-k showed hydrogen bond donor

CDK2 Protein Kinase Inhibitory Activity of 4a-k
The promising cytotoxicity of 4, especially 4j and 4k, motivated us to study further inhibitory activities against CDK2 protein kinase. As summarized in Table 2, 4j and 4k exhibited good inhibitory activity against cyclin-dependent kinase 2 (CDK2), which is half of that of the known kinase inhibitor imatinib.

In Silico Drug Likeness Property Analysis
Rational drug designing is the most significant part in modern drug discovery approaches. In this regard, computational ADME (absorption, distribution, metabolism and excretion) analysis can help us select the best drug in terms of cost, time and efficiency. Applying computational chemistry tools, in vitro and in vivo ADME prediction is now much more convenient and it can aid pharmaceutical industries to screen thousands of compounds within a short time [53]. Here, synthesized compounds (4a-k) were screened for predicted ADME values and the results are summarized in Table 3. Since high molecular weight compounds are always less effective in terms of intestinal absorption [54,55], our designed and synthesized isatin-hydrazones' (4a-k) molecular weights were kept low, in between 263-328 Da. Compounds 4a-k showed hydrogen bond donor (HBD) values of 1, except 4i which had a HBD value of 2 (recommended value = <5), and hydrogen bond acceptor (HBA) values of 5, except 4i which had HBA value = 6.5, 4h with a HBA value = 5.75 and 4d with a HBA value = 5.5 (recommended value = <10). On the other hand, doxorubicin (Doxo) showed a HBD value of 5 and HBA value of 15, which indicates that synthesized isatin-hydrazones are superior to Doxo in respect to HBD and HBA values. A parameter was established in 2002 to check the bioavailability of a drug using octanol/water partition coefficient and solubility scoring (recommended values for octanol/water partition coefficient are −2 -6.5 and solubility scoring are −6.5 -0.5 mol/dm−3) [56]. The octanol/water partition coefficient for hydrazones 4a-k is in between 1.79-3.12 and solubility score is −3.39 -−4.35, respectively. Doxo showed a score within the reference values of −0.49 and −2.37, respectively. The hERG K + channel blockers are potentially toxic for the heart, thus the recommended range for predicted logIC 50 values for blockage of hERG K + channels (loghERG) is > −5 [57]. Intriguingly, 4a-c and 4e-k showed higher values for loghERG score (> −5.58-−5.91) than Doxo (−6.02), except 4d, which was similar to Doxo, which proved their (4a-k) toxicity to be lower than Doxo. The Caco-2 cell, considered as the reliable in vitro model to estimate oral drug absorption and transdermal delivery [58], was high (>1310) for all compounds except 4i (487). Interestingly, Doxo had a much lower value (2.29) than 4a-k, which signifies the improved oral drug absorption and transdermal delivery efficiency of the studied compounds compared to Doxo.  <500); b hydrogen bond donor (acceptable range: ≤5); c hydrogen bond acceptor (acceptable range: ≤10); d predicted octanol/water partition coefficient (acceptable range: −2-6.5); e predicted aqueous solubility, S in mol/dm−3 (acceptable range: −6.5-0.5); f predicted IC 50 value for blockage of hERG K+ channels (concern: below −5); g Caco−2 value, permeability to Caco−2 (human colorectal carcinoma) cells in vitro; h blood−brain barrier permeability (acceptable range:~−0.4); i predicted apparent Madin-Darby canine kidney (MDCK) cell permeability in nm/sec, QPPMDCK= >500 is great, <25 is poor; j predicted human oral absorption on 0% to 100% scale (<25% is poor and >80% is high); k Doxo = Doxorubicin.
The blood-brain barrier separates the CNS from blood, and a successful compound must pass into the blood stream, which depends on several factors, such as molecular weight, which must be below 480 [59]. Since our synthesized compounds have low molecular weights and fall within the recommended values, this, therefore, showed significant results. Madin-Darby canine kidney (MDCK) cell permeability is considered as the measurement of blood-brain barrier permeability, where greater than 500 is of great value and less than 25 indicates a very poor result according to Jorgensen's rule of 3 [60]. Except compound 4i (227), all the other compounds gave much higher MDCK values (> 666 to 3162) than Doxo (0.766 only). The synthesized compounds also gave a predicted human oral absorption rate of 100%, except compound 4i which gave 86%. On the other hand, Doxo showed a predicted human oral absorption rate of 0%. Taken together, all the designed compounds, 4a-k, of this study showed higher predicted ADME values than Doxo.

Architecture of the CDK2 Active Site
Developing new inhibitors against CDK2 mainly involves designing compounds that can act as ATP competitive inhibitors by binding to the ATP binding cleft of CDK2. According to the active and inactive state of the protein kinase, two different types of inhibitors can be designed: type I and type II inhibitors. Type I inhibitors mainly bind to the ATP binding pocket of an active kinase, whereas type II inhibitors bind to the inactive kinase [61]. From the recently published crystal structure of CDK2 in a complex with the inhibitor CVT-313, it was found that active kinase inhibition depends solely on the interaction with the DFG motif, which comprises Asp145-Phe146-Gly147. Leu83, Asp86 and Asp145 form the ATP binding site of CDK2 through hydrogen bonds, where Asp145 belongs to the DFG motif. Outside of the active site, the residues Glu81-Leu83 hinge linker sequence is responsible for flexibility of the kinase. The phosphorylation of the C-terminal domain contains the catalytic residue (Glu51) required for the phosphorylation of Thr160 in the T-loop for its activation.
The activation segment is composed of the conserved DFG motif (Asp145-Phe146-Gly147) and the APE motif (Ala170-Pro171-Glu172). The unique PSTAIRE motif (Pro45-Glu51) in CDK2 that has a key role in its interaction with the cyclin subunit is found in the N-terminal domain [61]. To investigate whether the synthesized compounds (4j and 4k, based on best cytotoxicity assay and enzyme inhibition assay against CDK2 protein kinase) are type I or type II inhibitors, and also to check their binding mechanism with CDK2, we performed a molecular docking analysis.

In Silico Binding Mechanism Analysis
From the docking analysis of compounds 4j and 4k with CDK2, it was clearly observed that both compounds interacted with the ATP binding pocket residues Leu83 and Asp86 but not with Asp145 of the DFG motif ( Figure 4A,D). 4j and 4k thus can act as ATP competitive type II inhibitor by binding to inactive kinase [61]. Moreover, both the compounds showed a similar fashion of interactions, which involved several hydrogen bonds and ionic interactions with the binding site cavity surrounding residues, such as Ile10, Val18, Ala31, Val64, Glu81, Phe82, Leu134 and Ala144, which is consistent with the molecular docking analysis of 3,6-disubstituted pyridazines; 6-N,6-N-dimethyl-9-(2-phenylethyl)purine-2,6-diamine as CDK2 inhibitors (2, 3) [62,63]. However, compound 4j formed one additional interaction with Lys89, which was absent in case of compound 4k. It may bind to inactive kinase, which can be analyzed by finding no interaction between the compounds with catalytic residue Glu51 that is responsible for the phosphorylation of Thr160 for activation of kinase function. The interacting residues, although they do not belongs to the ATP binding pocket, formed a pathway for the compounds to bind properly to the ATP binding pocket. Glu81 to Leu83, on the other hand, forms the hinge region responsible for the flexibility of the protein kinase. From Figure 4C,F, it is visible how these residues form the binding cleft and pathway for the compounds to occupy the ATP binding site of CDK2 protein kinase. Figure 4B,E show the 3D interaction pattern and formation of binding cleft. From the docking analysis, it can be concluded that both 4j and 4k served as ATP competitive type II inhibitors by interacting with ATP binding pocket residues and with the residues that paved the way for the compounds to bind to the CDK2 ATP binding pocket. Table 4 summarizes the compounds and names the interacting residues, along with the types of interactions and the docking score of each compound.

General
Chemicals and solvents were of commercial reagent grade (Sigma-Aldrich, St. Louis, MO, USA) and were used without further purification. The progress of reactions and purity of reactants and products were checked using pre-coated silica gel 60 aluminum TLC sheets with fluorescent indicator UV254 of Macherey-Nagel, and detection was carried out with an ultraviolet light (254 nm) (Merck, Darmstadt, Germany). Melting points were measured using an Electrothermal IA9100 melting point apparatus (Stone, Stafforshire, ST15 OSA, UK). Infrared (IR) spectra (as KBr pellet) were recorded on a FT-IR Spectrum BX device from Perkin Elmer (Ayer Rajah Crescent, Singapore). 1 H NMR spectra were recorded using a Bruker 600 MHz spectrometer (Reinstetten, Germany) and DMSO-d 6 was used as a solvent. Chemical shifts were expressed in parts per million (ppm) relative to TMS as an internal standard. Mass spectra were taken with an Agilent 6410 Triple Quad mass spectrometer fitted with an electrospray ionization (ESI) ion source (Agilent Technologies, Palo Alto, CA, USA).

General Procedure for the Synthesis of 3-[benzylidene(substituted)hydrazono]indolin-2-ones 4a-k
A mixture of isatin monohydrazone (2, 5 mmol) and 4-methylbenzaldehyde (3a, 5 mmol) in absolute ethanol (15 mL) was added to a few drops of glacial acetic acid. The reaction mixture was refluxed for 4 h. The completion of the reaction was monitored by TLC. The precipitate solid was filtered, washed with cold ethanol and air dried, and was then further purified by recrystallization using ethanol, obtained 4a as yellow powder. Please see in Supplementary Materials for NMR ( 1 H & 13 C) and MS spectra of compound 4 (in Supplementary Materials).  13

Cytotoxicity
The cytotoxicity of the synthesized compounds was evaluated by MTT assay, as previously described [66]. Two cancer cell lines, MCF7 (human breast adenocarcinoma) and A2780 (human ovary adenocarcinoma), were used in this study, which were obtained from the ATCC (Rockville, MD, USA). They were sub-cultured in RPMI-1640 media (supplemented with 10% FBS and 1% antibiotics) at 37 • C and 5% CO 2 . Additionally, compounds were prepared at the same medium to obtain serial dilutions (50,25,19,1 and 0.1 µM). The two cell lines were separately cultured in 96-well plates (3 × 10 3 /well) and incubated at 37 • C overnight. The following day, before treating the cells with the compounds, each well of the T0 plate was treated with 50 µL MTT solution (2 mg/mL in phosphate buffered saline) and then incubated for 2-4 h. The media were aspirated, and the formazan crystals were solubilized by adding 150 µL DMSO. Absorbance was read on a multi-plate reader (BioRad) at 550 mm. Optical density of the purple formazan A550 was proportional to the number of viable cells. Compound concentration causing 50% inhibition (IC 50 ) compared to control cell growth (100%) was determined. The data were obtained from triplicates and analyzed using statistical software.

In Vitro Cyclin Dependent Kinase2 (CDK2) Inhibitory Activity
The CDK2 Assay Kit is designed to measure CDK2/CyclinA2 activity for screening and profiling applications, using Kinase-Glo ® MAX as a detection reagent. The CDK2 Assay Kit comes in a convenient 96-well format, with enough purified recombinant CDK2/CyclinA2 enzyme, CDK substrate peptide, ATP and kinase assay buffer for 100 enzyme reactions [67]. The assay was performed according to the protocol supplied from the CDK2 Assay kit #79599. The CDK2/CyclinA2 activity at a single dose concentration of 10µM was performed, where the Kinase-Glo MAX luminescence kinase assay kit (Promega#V6071) was used. The compounds were diluted in 10% DMSO and 5 µL of the dilution was added to a 50 µL reaction so that the final concentration of DMSO was 1% in all of the reactions. All of the enzymatic reactions were conducted at 30 • C for 40 min. The 50 µL reaction mixture contained 40 mM Tris, pH 7.4, 10 mM MgCl 2 , 0.1 mg/mL BSA, 1 mM DTT, 10 mM ATP, Kinase substrate and the enzyme (CDK2/CyclinA2). After the enzymatic reaction, 50 µL of Kinase-Glo ® MAX Luminescence kinase assay solution was added to each reaction and the plates were incubated for 5 min at room temperature. Luminescence signal was measured using a Bio Tek Synergy 2 microplate reader.

Molecular Docking and In-Silico ADME Analysis
For molecular docking purposes, the Protein Data Bank (PDB) structure corresponding to the CDK2 protein kinase was downloaded from the Research Collaboratory for Structural Bioinformatics (RCSB) PDB database (https://www.rcsb.org/) in PDB format. The PDB ID used for CDK2 protein kinase was 2BHY. Proteins and compounds were prepared for docking by using an established procedure [68]. Discovery Studio was used for making 2D interaction figures. Pymol was used to generate the 3D and surface representation figures. For the in silico ADME analysis, all the compounds' structures were prepared with the LigPrep module of Schrodinger Maestro and ADME was calculated by the Qikprop module of the same software package [69].

Conclusions
A series of novel isatin-hydrazones (4a-b and 4d-k), with a known compound 4c, were designed and synthesized with good to moderate yields for cytotoxicity evaluation for the development of potent anticancer therapeutics. Among the compounds, 4j showed a two-fold increase in cytotoxicity compared to the known cancer drug doxorubicin, and 4k showed a similar cytotoxicity. The IC 50 value of compound 4j was 1.51 and for 4k it was 3.56 µM, whereas doxorubicin had a 3.1 µM concentration against human breast adenocarcinoma (MCF7) cell lines. The most active compounds, 4j and 4k, were further evaluated for their inhibitory activities against CDK2 protein kinase. As expected, 4j and 4k exhibited good inhibitory activity against cyclin-dependent kinase 2 (CDK2) 0.2456 and 0.3006 µM, respectively, which is comparable to kinase imatinib 0.1512 µM. Highly recommended predicted ADME values were obtained than the known doxorubicin. The molecular docking study of 4j and 4k with CDK2 protein kinase revealed that they interacted with ATP binding pocket residues and lacked interactions with the active state DFG motif residues; therefore, 4j and 4k can be considered as ATP competitive type II inhibitors against CDK2 protein kinase. In conclusion, these simple molecules, isatin-hydrazones 4j and 4k, can be used as potential agents for anticancer therapeutics in further mechanism and toxicity studies.