Synthesis, Cytotoxic Evaluation, and Structure-Activity Relationship of Substituted Quinazolinones as Cyclin-Dependent Kinase 9 Inhibitors

Cyclin-dependent kinase 9 (CDK9) plays a critical role in transcriptional elongation, through which short-lived antiapoptotic proteins are overexpressed and make cancer cells resistant to apoptosis. Therefore, CDK9 inhibition depletes antiapoptotic proteins, which in turn leads to the reinstatement of apoptosis in cancer cells. Twenty-seven compounds were synthesized, and their CDK9 inhibitory and cytotoxic activities were evaluated. Compounds 7, 9, and 25 were the most potent CDK9 inhibitors, with IC50 values of 0.115, 0.131, and 0.142 μM, respectively. The binding modes of these molecules were studied via molecular docking, which shows that they occupy the adenosine triphosphate binding site of CDK9. Of these three molecules, compound 25 shows good drug-like properties, as it does not violate Lipinski’s rule of five. In addition, this molecule shows promising ligand and lipophilic efficiency values and is an ideal candidate for further optimization.


Structure-Activity Relationship (SAR) Analysis
The CDK9 inhibition activities of the target compounds were evaluated, with flavopiridol used as a reference. Compound 1 has a similar scaffold to those of several reported CDK9 inhibitors and Scheme 2. Synthesis routes of compounds 23-28.

Structure-Activity Relationship (SAR) Analysis
The CDK9 inhibition activities of the target compounds were evaluated, with flavopiridol used as a reference. Compound 1 has a similar scaffold to those of several reported CDK9 inhibitors and shows promising CDK9 inhibitory activity with an IC 50 value of 0.644 µM [39]. Therefore, it was selected as a hit compound and modified with different substituents at the quinazoline position 2 to improve its inhibition properties. The results show that introducing acetamide and acetanilide groups at position 2 in compounds 2 and 3 showed better inhibition than the hit compound, with IC 50 values of 0.454 and 0.421 µM, respectively, as shown in Table 1. To investigate the effects that other functional groups substituted at the p-position of the acetanilide ring have on improving the potency of the compound, analogs 4−8 with various substituents, such as methyl, acetyl, methoxy, and ethoxy groups, were prepared and evaluated. Compound 4, with a p-methyl group, displays lower inhibitory activity than compounds (1-3), with an IC 50 value of 0.788 µM, as shown in Table 1. shows promising CDK9 inhibitory activity with an IC50 value of 0.644 μM [39]. Therefore, it selected as a hit compound and modified with different substituents at the quinazoline pos to improve its inhibition properties. The results show that introducing acetamide and acetanilid groups at position 2 in compounds 2 and 3 showed better inhibition than the hit compound, wit values of 0.454 and 0.421 μM, respectively, as shown in Table 1. To investigate the effects that oth functional groups substituted at the p-position of the acetanilide ring have on improving the pot the compound, analogs 4−8 with various substituents, such as methyl, acetyl, methoxy, and e groups, were prepared and evaluated. Compound 4, with a p-methyl group, displays lower itory activity than compounds (1-3), with an IC50 value of 0.788 μM, as shown in Table 1. Furthermore, substituting the p-methyl group for a larger group, such as an group used to generate compound 5, was found to be detrimental as it was less than the previous analogs, showing an IC50 value of 0.829 μM, as shown in Table 1 ever, introducing a methoxy group at the p-position gave compound 6, which le recovery in the inhibitory activity, exhibiting an IC50 value of 0.463 μM. Interes when an ethoxy group was introduced, as in compound 7, the inhibitory activity inc by around 6-fold compared to the hit compound 1, with the compound exhibiting 6-fold compared to the hit compound 1, with the compound exhibiting an IC 50 value of 0.115 µM. However, replacing the p-methoxy group with 3,4,5-trimethoxy groups in the acetanilide ring led to compound 8, which demonstrated less potency than the hit compound, exhibiting an IC 50 value of 0.501 µM. Interestingly, introducing a halogen (such as Br, Cl, or F) at the p-position of acetanilide led to the generation of compounds 9-11, which show excellent inhibition of CDK9. Furthermore, introducing bromine at this position, as in the formation of compound 9, seemed to be more favorable to the inhibitory activity compared with the introduction of other halogens at the same position, with compound 9 showing an IC 50 value of 0.131 µM. Analogously, increasing the lipophilicity of the benzylacetamide substituents in compounds 12-14 could be essential for improving their CDK9 inhibitory activity compared to their corresponding acetanilide-containing counterparts, such as compounds 6 and 11. It was observed that the phenylpropanamide substituents in compounds 15 and 16, which exhibit

Molecular Docking
Molecular docking experiments were performed using the genetic algorithm docking program GOLD 5.2 to rationalize the observed potency of compounds 7, 9, and 25. In addition, flavopiridol was used as a reference compound to compare the binding pattern. The modeled complexes with CDK9 are shown in Figure 4.
At the binding cavity of CDK9, flavopiridol, compounds 7, 9, and 25 occupy the adenosine triphosphate (ATP) binding site. Compounds 7 and 9 display similar binding modes. These conformations, however, are different from the ones shown by flavopiridol. The benzene ring of the quinazolinone is in contact with the Phe105 residue in the hinge region. In addition, the sulphonamide moieties of compounds 7 and 9 bind differently, forming hydrogen bonds with the Thr29 and Glu107 residues in the enzyme, respectively. Moreover, the Ethoxy group of compound 7 forms a hydrogen bond with Lys48, whereas there is no interaction between the enzyme and the anilide moiety of compound 9. This suggests that the sulphonamide moiety is essential for the activity, whereas the acetanilide groups could be necessary for the potency in the context of 2-mercapto-3-(4-sulfamoylphenethyl)quinazolin-4-one. However, compound 25 adopts a similar orientation to that of flavopiridol. Unlike compounds 7 and 9, the NH group of compound 25 forms a hydrogen bond with Glu107. Moreover, the methyl group of compound 25 is in contact with Phe103 of the gatekeeper region of the enzyme. In addition, the m-bromophenyl group in compound 25 occupies a similar position to that occupied by the o-chlorophenyl ring of flavopiridol. Breast cancer often occurs with dysregulation in CDK9 levels, and several studies have shown the efficacy of CDK9 inhibitors in breast cancer [20,21]. Therefore, the antiproliferative activity of compounds 1-20 and 23-29 was evaluated against the breast cancer cell line MCF-7 by the metabolic assay MTT. Compounds 1-16 showed potent cytotoxic activities with IC 50 values ranging from 0.16 to 4.65 µM, with compounds 4 and 5 being the most potent cytotoxic agents, which could be due to their multitarget inhibitory activities against EGFR, HER2, and VEGFR2 as well as CDK9 [26]. However, compounds 17-20 and 23-29 were significantly less potent than compounds 1-16, with IC 50 values ranging from 3.88 to 28.7 µM.

Molecular Docking
Molecular docking experiments were performed using the genetic algorithm docking program GOLD 5.2 to rationalize the observed potency of compounds 7, 9, and 25. In addition, flavopiridol was used as a reference compound to compare the binding pattern. The modeled complexes with CDK9 are shown in Figure 4.
At the binding cavity of CDK9, flavopiridol, compounds 7, 9, and 25 occupy the adenosine triphosphate (ATP) binding site. Compounds 7 and 9 display similar binding modes. These conformations, however, are different from the ones shown by flavopiridol. The benzene ring of the quinazolinone is in contact with the Phe105 residue in the hinge region. In addition, the sulphonamide moieties of compounds 7 and 9 bind differently, forming hydrogen bonds with the Thr29 and Glu107 residues in the enzyme, respectively.

Lipinski's Rule of Five
Lipinski's rule of five was used to evaluate the drug-like properties of compounds 1-20 and 23-29. DataWarrior was used to estimate the molecular weight (MW), CLogP, hydrogen bond acceptors (HBAs), and hydrogen bond donors (HBD) for each molecule, and the values are presented in Table 3. The data in the table show that compounds 4-16 have molecular weights of >500 Da. In addition, compound 8 offers an additional violation of Lipinski's rule, with HBAs >10. However, all the compounds satisfy Lipinski's rule regarding lipophilicity and the number of HBDs, with CLogP and HBD values below 5.  Moreover, the Ethoxy group of compound 7 forms a hydrogen bond with Lys48, whereas there is no interaction between the enzyme and the anilide moiety of compound 9. This suggests that the sulphonamide moiety is essential for the activity, whereas the acetanilide groups could be necessary for the potency in the context of 2-mercapto-3-(4sulfamoylphenethyl)quinazolin-4-one. However, compound 25 adopts a similar orientation to that of flavopiridol. Unlike compounds 7 and 9, the NH group of compound 25 forms a hydrogen bond with Glu107. Moreover, the methyl group of compound 25 is in contact with Phe103 of the gatekeeper region of the enzyme. In addition, the m-bromophenyl group in compound 25 occupies a similar position to that occupied by the o-chlorophenyl ring of flavopiridol.  Table 3. The data in the table show that compounds 4-16 have molecular weights of >500 Da. In addition, compound 8 offers an additional violation of Lipinski's rule, with HBAs >10. However, all the compounds satisfy Lipinski's rule regarding lipophilicity and the number of HBDs, with CLogP and HBD values below 5. The LE is a property that describes the potency per heavy atom of a drug [50][51][52]. The LE values of the synthesized compounds were obtained using DataWarrior according to the following equation [50,52]: N represents the number of heavy atoms, i.e., non-hydrogen atoms in the drug, R is the universal gas constant, T is the absolute temperature in degrees Kelvin, and IC 50 is CDK9 IC 50 in mol/L.
LE is an essential metric in lead optimization, which allows the comparison of the affinity of molecules according to their size. Compounds with LE values higher than 0.3 are considered promising lead compounds. The LE values for the target compounds in this study are presented in Table 4, which shows that the LE values of the synthesized compounds are between 0.2 and 0.7. Except for compounds 3-16, the LE values fall into an acceptable range and are >0.3 [52,53]. LLE is used to link the potency of a compound to its lipophilicity [52,53]. The challenge in drug discovery is optimizing a compound's activity while maintaining lipophilicity at a constant value. For this reason, LLE is considered an effective strategy to control the lipophilicity of a molecule to avoid any "molecular obesity" during lead optimization. The LLE values for compounds 1-20 and 23-29 shown in Table 4 were obtained using DataWarrior according to the following equation [52]: pIC 50 is the negative log of the CDK9 IC 50 and CLogP is the calculated LogP value. An acceptable lead compound should have an LLE value of ≥ 5. Compounds 2, 17, and 27 show good LLE values, i.e., LLE > 5 [52,53]. However, the other compounds have values that are below the recommended limit.
In conclusion, compound 17 is a good candidate for lead optimization since it has the lowest non-hydrogen atoms (N), an acceptable LE value of 0.708, and an acceptable LLE value of 5.20.

CDK9 Kinase Assay
In vitro luminescent CDK9 kinase assay was performed as reported previously using Kinase-Glo ® MAX as a detection reagent [43]. Briefly, 5 µL of each inhibitor in concentrations ranging from 10 µM to 1 nM (10 µM, 1 µM, 0.1 µM, 0.01 µM, and 0.001 µM) and 10 µL of enzyme substrate were mixed in 20 µL of kinase assay buffer (obtained from BPS Bioscience, catalog #79334) at room temperature. Then 20 µL of 5 ng/µL CDK9/cyclin T was added to the mixture to initiate the reaction. After 45 min, 50 µL of Kinase-Glo ® Max reagent was added, and the resulting mixture was incubated for 15 min at room temperature. The chemiluminescence was measured microplate reader, and IC 50 values were calculated using Prism 8.0 (GraphPad Software, San Diego, CA, USA).

Molecular Docking
Molecular docking was performed according to the procedure reported previously using the X-ray crystal structure of flavopiridol in a complex with CDK9 (PDB ID: 3BLR) which was retrieved from the PDB Data Bank (URL: http://www.rcsb.org; accessed on 20 September 2022) [26].

Conclusions
Twenty-seven compounds were synthesized, and their CDK9 inhibitory and cytotoxic activities were evaluated. Compounds 7, 9, and 25 were the most potent CDK9 inhibitors, with IC 50 values of 0.115, 0.131, and 0.142 µM, respectively. The binding modes of these molecules were studied using molecular docking, which showed that they occupy the ATP binding site of CDK9. Of these three molecules, compound 25 shows good drug-like properties since it does not violate Lipinski's rule of five. In addition, this molecule shows promising LE and LLE values and is an ideal candidate for further optimization.