Novel 2,4-disubstituted quinazoline analogs as antibacterial agents with improved cytotoxicity profile: Modification of the benzenoid part

Bacterial resistance to currently used antibiotics demands the development of novel antibacterial agents with good safety margins and sufficient efficacy against multi-drug resistant isolates. We have previously described the synthesis of N -butyl-2-(butylthio)quinazolin-4-amine ( I ) as an optimized hit with broad-spectrum antibacterial activity and low cytotoxicity. In addition, we have identified a potential growing vector for this series of compounds. Herein, we describe further hit optimization which includes systematic diversifications of both the benzenoid part and the substituents at position 6 and 7 of compound I . Growing of the molecule beside the core modifications yielded several compounds with remarkable anti(myco)bacterial activity against a panel of pathogenic bacteria, including drug-resistant strains. Compound 12 showed a 2-4 fold improvement in activity than I against S. aureus Newman , S. pneumoniae DSM-20566 and E. faecalis DSM-20478 . The compounds also showed a good safety profile towards human HepG2 cells.

The inappropriate and excessive use of antibiotics, without medical supervision, led to the increase and the emergence of lethal human infections that are resistant to multiple antibiotics.2][3][4][5] Staphylococci and enterococci are Gram-positive bacterial pathogens that are the causative agents for several hospital-and community-acquired infectious diseases.Staphylococcus aureus causes different types of infections, ranging from simple infections in skin and soft tissues to more serious diseases such as infective endocarditis, sepsis and pneumonia. 6, 7The spread of methicillin-resistant S. aureus (MRSA) is on escalation and presents a major threat.
][10][11] Furthermore, the enterococcal species Enterococcus faecium and Enterococcus faecalis are the most common enterococcal species cultured from patients, accounting for more than 90% of clinical enterococcal isolates. Streptococcus pneumoniae is a Gram-positive bacterial pathogen and a causative agent of respiratory tract infections such as meningitis, sinusitis, pneumonia and acute otitis media.It produces numerous virulence factors involved in the progress of the disease. 16Several reported cases of endocarditis were caused by penicillin-resistant S. pneumoniae (PRSP) strains. 17Over the past years, S. pneumoniae has developed resistance to several classes of antibiotics, including macrolides, beta-lactams, fluoroquinolones, lincosamides, tetracyclines and trimethoprim-sulfamethoxazole. 18nsequently, the need for new antibacterial drugs is now of paramount importance.
Quinazoline is a well-known scaffold exhibiting a wide range of different biological activities including anticancer, 6, 19 anti-inflammatory 20-22 and antimicrobial activities. 23, 244][35] A series of 2-substituted quinazoline derivatives with broad-spectrum antibacterial activity through the inhibition of the transcription/translation in different bacterial species was previously reported. 36In an earlier study, Harris et al. developed a number of 5substituted-2,4-diaminoquinazolines that revealed inhibitory activity against the bacterial dihydrofolate reductase (DHFR) enzyme in S. aureus and in Escherichia coli.However, this series of synthesized compounds lacked specificity, as the compounds did not only inhibit bacterial DHFR but they were also active against its bovine liver counterpart.
Optimization of the substituents on the quinazoline resulted in higher selectivity towards the bacterial enzyme. 37In another recently published study, some quinazolin-4-ones were also reported as cell wall biosynthesis inhibitors, through their ability to bind to the penicillin-binding protein 22 (PBP)2a. 38Another series of N 2 ,N 4 -disubstituted quinazoline-2,4diamine derivatives was reported in the literature to have antibacterial activity against MRSA. 39cently, we reported the synthesis and antibacterial activity of an optimized hit compound I (Figure 1) with promising antibacterial activity against Gram-positive bacteria, including drug-resistant strains, the Gram-negative bacterium E. coli TolC as well as Mycobacterium smegmatis, with MIC values ranging from 2-8 µg/mL. 40Importantly, compound I showed also a good safety profile rationalized by the decreased toxicity against A549 and HepG2 cells, in addition to the lack of hemolytic activity at concentrations up to 500 µM.Noteworthy, its sulfur analog (II) was inactive.
We also identified position 6 as a potential growing vector for this series of compounds.Herein, we describe further structure-activity relationship (SAR) exploration and optimization of compound I while maintaining a safe cytotoxicity profile (Figure 2).The modifications adopted include variable substitutions at positions 6 and 7 (III).These substituents include halogens, which are known to efficiently influence various properties such as membrane permeability, intramolecular interactions, pharmacokinetic properties and others. 41, 42Additionally, analogs with different aryl and heteroaryl substitutions at position 6 or 7 were also prepared.Another group of compounds was prepared by replacing the benzenoid part of the quinazoline ring with the two isosteric heterocycles -thiophene (IV) and pyrazole (V) -to give thieno [2,3-d]pyrimidine and pyrazolo [3,4-d]pyrimidine scaffolds.Finally, we further derivatized compound (II) at its position 6 (VI) in a trial to recover its antibacterial potency.The antibacterial activity of all newly synthesized target compounds was tested in vitro against both drug-sensitive and drug-resistant bacteria.
The preparation of the quinazoline-2,4-dithione analogs 2a-f was achieved by the reaction of the orthoaminobenzonitrile derivatives 1a-f with carbon disulfide in pyridine at 70 °C.The 2,4-bis(butylthio)-quinazoline derivatives 3a-f were obtained by the alkylation of compounds 2a-f with 1-bromobutane, in the presence of K2CO3.The obtained 2,4-bis(butylthio)-quinazoline derivatives 3a-f were then reacted with n-butylamine which underwent a regioselective substitution at position C4 43-45 to afford the 4-N-butyl derivates 4a-f.(Scheme 1).In order to further explore and optimize the hit compound I while maintaining a safe cytotoxicity profile, several modifications were adopted.These modifications included variable substitutions at positions 6 or 7 with either halogens or further extending with aryl or hetero aryl systems.Besides, replacing the benzenoid part of the quinazoline with other heterocycles.
The insertion of halogens (either fluorine, chlorine, or bromine) at position 6 or 7 on the quinazoline scaffold of compound I (4a-f) diminished the antibacterial activity.The electron-withdrawing effect of the halogens was detrimental since all compounds exhibited MIC values higher than 128 µg/mL against the tested strains.
The second group of compounds were those substituted at positions 6 or 7 with aryl or heteroaryl rings ( 5-16).Most of the selected aryl/heteroaryl rings had polar features to minimize the gain in lipophilicity and associated solubility problems.The plain phenyl at position 6 or 7 (compounds 7 and 5, respectively) abolished the antibacterial activity.
Next, we tried to introduce a carboxylic acid group at meta-and para-position of the phenyl ring (compounds 8 and 9, respectively).Although the p-COOH derivative (8) was completely inactive, shifting of the -COOH to the meta position (9) dramatically revived the activity against the tested Gram-positive bacteria to give one of the most potent compounds in the present series.Compound 9 showed remarkable antibacterial activity against all tested Gram-positive bacteria, including the drug-resistant S. aureus and S. pneumoniae strains, with MIC values ranging from 4-16 µg/mL.However, compared to the initial frontrunner compound I, compound 9 did not show activity against M. smegmatis mc 2 155.
Replacing the m-COOH with the less acidic phenolic OH (compound 12) slightly enhanced the potency in comparison with compound 9, with compound 12 displaying MIC values of 1-4 µg/mL against S. aureus, S. pneumonia, and E.
faecalis, and it also showed activity against M. smegmatis (MIC 16 µg/mL).Unfortunately, although being active against drug-sensitive strains and PRSP, 12 was inactive against the MRSA/VISA strains.Neither the alcoholic OH (compound 10) nor the cyano function (compound 11) succeeded to replace the phenolic group while maintaining the potency.
Moreover, using the meta-methoxy phenyl (compounds 13) was also not beneficial.Replacing the phenyl in compound 7 with the 3-or 4-pyridyl (14 and 15, respectively) or 4-pyrazolyl ( 16) failed to regain the high potency of compound I.In conclusion, the acidic function (m-COOH and m-OH) in the 6-aryl extension on the quinazoline ring seemed a strict requirement for high potency in this cluster of compounds.
Replacing the benzenoid part of the quinazoline compound I with a thiophene gave the thieno [2,3-d]pyrimidine analogue 20, which displayed good antibacterial activity against S. aureus Newman strain, S. pneumoniae and the effluxdeficient E. coli strain with MIC in the one-digit µg/mL range.However, activity against Enterococcus spp.and VISA was completely abolished.
The benzenoid part was also replaced with the bioisostere pyrazole represented in compounds 25a-c and 26.This modification in general did not seem to be beneficial, with compounds 25a, 25b and 26 showing only moderate antibacterial activity against selected strains of the test panel, while 25c with n-butyl substitution at position 2 was inactive against all tested strains.Shortening the S-butyl of 25b into methyl as in compound 25a decreased the antibacterial activity against E. coli TolC (MIC = 64 µg/mL for 25a vs. 4 µg/mL for 25b), while marginally improved the antibacterial activity against the three strains of S. aureus (MIC = 64 µg/mL for 25a and >128 µg/mL for 25b), including the drug-resistant strains.Moving the n-butyl chain to position 1 (26) led to a virtually inactive compound but 26 unexpectedly inhibited the growth of S. pneumoniae with an MIC of 8 µg/mL.
To further explore the antibacterial activity of compound II, we investigated the effect of adding some polar functional groups like amino, sulfonamide and urea at position 6 of the quinazoline nucleus.The presence of an amino group at position 6 (30) gave one of the most active compounds in this study with similar potency to compound I, but it lacked activity against E. coli K12 ΔtolC.Compound 30 showed a broad antibacterial activity against the tested Grampositive bacteria, including the drug-resistant strains, as well as M. smegmatis (MIC values ranging from 4-32 µg/mL).
According to the SAR of the synthesized compounds, it was found that the presence of an acidic functional group on a phenyl ring at position 6 of the quinazoline ring is beneficial for the antibacterial activity.In addition, replacing the benzenoid part of the quinazoline ring with thiophene or pyrazole rings led to the discovery of new potential antibacterial thieno [2,3-d]pyrimidine and pyrazolo [3,4-d]pyrimidine analogs with the thieno [2,3-d]pyrimidine being more potent.It was also found that the insertion of a (substituted-)amino group at position 6 of the inactive 2,4-di-thiobutyl quinazoline could help to regain antibacterial activity of the analogs (30 and 32), Interestingly, the quinazoline derivatives 9, 30 and 32 and the thieno [2,3-d]pyrimidine derivative 20 showed no cytotoxicity in HepG2 cells.Only compound 12 displayed moderate cytotoxicity at a concentration higher than its MICs against four of the tested bacterial strains.Overall, the applied modification approaches, molecule growing and scaffold replacement were successful to produce novel promising antibacterial quinazoline analogs with a favorable selectivity index.

Figure 1 :
Figure 1: Optimized hit compound I showing the potential growing position and its inactive analog compound II

Table 1 (
Table S1 in Supplementary Information is showing the data in µM concentration unit).

Table 1 :
MIC values of the synthesized analogs.

Table 2 :
Growth inhibitory activity against HepG2 cells Results are from two independent experiments and IC50 is given as average value (SD <10%).