Synthesis, in vitro Cytotoxicity and Trypanocidal Evaluation of Novel 1,3,6-Substituted Non-fluoroquinolones

Sleeping sickness (trypanosomiasis) is a neglected tropical disease that affects mostly the poorest communities in sub-Saharan Africa. Toxic side effects associated with the use of current anti-trypanosomal drugs, which in some cases kill faster than the disease itself, necessitate the search for new drugs with better safety margins. To this effect, a small library bearing different substituents at position -1, -3, and -6 of the quinolone nucleus were synthesized and evaluated in vitro against HeLa cell lines and Trypanosoma brucei brucei for cytotoxicity and trypanocidal potentials, respectively. While most of these compounds showed no cytotoxic effect, they exhibited moderate to weak anti-trypanosomal activities. The SAR studies of this series provide new information worth considering in future exploration of the quinolone scaffold in search of more potent and safe trypanocidal agents.


Introduction
Human African trypanosomiasis (HAT), also referred to as sleeping sickness, 1 is among the WHO's list of neglected tropical diseases (NTDs). 2 It mostly affects people living in rural areas of sub-Saharan Africa, 3 where medical facilities are scarce and drug purchasing power of inhabitants is very low. 4 At least 2184 new cases of the disease were reported in 2016, 5 and approximately 60 million people living in 36 different countries are presently at risk of contracting the disease. 6 HAT takes two forms caused by two different subspecies of Trypanosoma brucei. 7 T. b. gambiense is the pathogenic subspecies causing the form of HAT commonly found in central and western Africa, 8,9 whereas T. b. rhodesiense is the subspecies responsible for the form of HAT prevalent in eastern and southern Africa. 10 It has been noted that the two sub-pathogenic species coexist in Uganda. 11 These pathogens are transmitted between humans following the bite of an infected tsetse fly. 12 The disease exists in two stages: 13,14 haemolymphatic stage -wherein the parasites are localized in blood and lymphatic systems, 15 and an encephalitic stage -wherein the parasites have invaded the central nervous system. 16,17 Current treatment options are limited to just four drugs. 18 In any case of HAT, the drug to be used is dictated by the form and stage of the disease. Pentamidine is the drug of choice for treating the haemolymphatic stage of HAT caused by T. b. gambiense, 19 while a combination of nifurtimox and eflornithine is used to treat the encephalitic stage. 20 In cases of T. b. rhodesiense, suramin is the recommended drug for treating the haemolymphatic stage, 21 while malarsoprol is used to treat the encephalitic stage. 22 Besides limited treatment options, the foregoing drugs are far from ideal. They all have poor oral bioavailability (and hence are administered intravenously), and considerable toxic side effects. 23 For example, pentamidine causes hyper or hypoglycaemia and hypotension, 24 suramin causes renal failure, eflornithine causes alopecia and seizures, 25 while at least 5.9 % of patients on malarsoprol die from its toxicity, 26 creating a scenario wherein patients either die from an acute illness or die faster from a pill.
The overwhelming life-threatening side effects of existing drugs used to treat sleeping sickness create a dire need for new compounds with better drug properties such as high oral bioavailability, and a high safety margin. The therapeutic potentials of quinolones cannot be over emphasized. Compounds containing this scaffold are currently in use as drugs to treat bacterial and viral infections as well as other conditions such as cancers. 27 Fluoroquinolones (Fig. 1a) in clinical use have been extensively screened against trypanosomes and their activity profiles established as moderate to poor. 28,29,30 A hit optimization study by Hiltensperger and co-workers generated a potent library of fluoroquinolones characterized by benzylamides, (a)cyclic amines, and aliphatic chains at positions -3, -7, and -1 of the quinolone nucleus. However, the lead compound (Fig. 1b) in this series suffers from poor solubility, 31 necessitating further work on this class of compounds. Unlike fluoroquinolones, the anti-trypanosomal potentials of non-fluorinated quinolones have not been extensively investigated, with just one study on non-fluorinated quinolones bearing substituents at position -1 and -2, respectively, being reported 32 (Fig. 1c). To further expand the SAR around the non-fluoquinolone scaffold, we conceptualized and synthesized a library of 18 non-fluorinated quinolones bearing unique concurrent substituents at position -1, -3, and -6. Target compounds were subjected to in vitro cytotoxicity and anti-trypanosomal evaluation. At 20 µM concentration, most of the compounds exhibited less than 50 % parasite viability while having little effect on the viability of HeLa cell lines (see Supplementary information). This suggests that the quinolone scaffold can be tailored via chemical synthesis to generate safe and potent drug substrates to treat trypanosomiasis.

Results and Discussion
We conceptualized and synthesized a set of 18 novel compounds that bear different substituents at positions -1, -3 and -6 of the quinolone scaffold. Conceptualized compounds were synthesized as depicted in Scheme 1. Briefly, 4'-nitroacetophenone was reduced to 4'-aminoacetophenone using reduced iron powder and acetic acid. 33 4'-Aminoacetophenone and 4-chloroaniline were each treated with diethyl ethoxymethylenemalonate in refluxing acetonitrile to form condensed methylenemalonate esters, which underwent cyclization in boiling diphenylether at 245-250°C for 5 min to form compound 2. Deprotonation of 2 using K 2 CO 3 , followed by N-alkylation with alkylhalides afforded target compounds 3a-h in yields ranging between 40 and 70 %. Ester functionality in compounds 3d-h underwent selective aminolysis in the presence of a ketone to afford target compounds 4a-j in 30-50 % yields. This transformation was realized using DBU as a base. All targeted compounds were characterized using proton and carbon NMR, HRMS and IR. The carbon NMR spectra of all compounds show a peak at c. 174 ppm, which is indicative of an oxo carbon within the quinolone nucleus (C-4) and the peak at c. 164 ppm is assigned to ester or amide carbonyl carbon (C-3a). With the exception of compounds 3a-c, the proton NMR spectra of all compounds show a singlet signal at~2.3-2.6 ppm, which is assigned to a methyl attached to a carbonyl carbon (CH 3 C=O). The carbonyl carbon is also evident in the respective carbon spectra at~197ppm. The presence of a triplet peak (J = 5.9 Hz) at c. 9.9 ppm in the proton NMR spectra of compounds 4a-j, which is absent in compounds 3d-h, suggests the successful conver-  sion of ester to amide. It is also worth noting that the acquisition of NMR spectroscopic data was at times hindered by compounds crystallizing out of the solutions. This sometimes necessitated the use of hot solvents to encourage compounds to remain in solution long enough to obtain 1H and 13C NMR spectroscopic data (see Supplementary information). The absorption band at 1654/cm on the IR spectrum further confirms the presence of amide. This focused library was screened in vitro against human cervix adenocarcinoma (HeLa) cell lines to investigate potential cytotoxicity effects. The compounds were incubated at 20 µM in 96-well plates containing HeLa cells for 48 h. The numbers of cells surviving upon drug exposure were determined using resazurin reduction to resorufin by live cells and reading resorufin fluorescence in a multiwell plate reader. Compounds were tested in duplicate, and a standard deviation (S.D.) calculated. Results are expressed as % viability based on fluorescence readings in treated wells versus untreated control wells. Emetine (which induces cell apoptosis) was used as positive control. With the exception of compound 4f (% viability, -19 %), and 4j (% viability, -13 %), which strongly inhibited HeLa cell lines, the rest of the series had little to no effect on HeLa cell viability. This observation suggests that this series with necessary optimization could serve as templates for the development of non-toxic anti-trypanosomal agents.
The compounds were further evaluated in vitro for antitrypanosomal activities by screening against the 427 strain of T. b. brucei. Compounds were added to in vitro cultures of T. b. brucei in 96-well plates at a 20 µM concentration. After an incubation period of 48 h, the numbers of parasites surviving drug exposure were determined by adding resazurin. As in HeLa cells, resazurin is reduced to resorufin by living parasites. Resorufin is a fluorophore (Exc 560 /Em 590 ) and can thus be quantified in a multiwell fluorescence plate reader. Compounds were tested in duplicate wells, and a standard deviation (S.D.) calculated. Results are expressed as % viability -the resorufin fluorescence in compound-treated wells relative to untreated controls. Pentamidine (an existing drug for treatment of trypanosomiasis) was used as a positive control. At 20 µM concentration, 11 compounds inhibited parasite growth below 50 % (Table 1); however, only compounds inhibiting parasite viability below 25 % with little or no effect on HeLa cell line were considered for IC 50 determination. The IC50 values for the selected compounds are summarized in Fig. 2.
Structure-activity relationship analysis across this series suggests that modifications at position -1, -3, and -6 of the quinolone scaffold influence anti-trypanosomal activities. For example, comparing the activity profiles of compounds 3b (IC 50 , 19 µM) and 3g (IC 50 , 24 µM), all having the same substituents at position -1 and -3 and differing in substitution pattern only at position -6 suggests that a chloride substituent at this position leads to increase anti-trypanosomal activity than a ketone. Also, comparing compounds 3d (IC 50 , 32 µM) bearing an alkyl chain at position -1 and 3g (IC 50 , 24 µM) bearing a substituted-benzyl moiety at position -1, indicates that the presence of a substi-   for compound 3a, 3b, 3d, 3g, 4b and 4g as well as the standard (PE = pentamidine) and their corresponding IC 50 values. tuted-benzyl moiety at position -1 seems to promote antitrypanosomal activities over alkyl chains. We also observed that the substituent on the benzyl moiety at position -1 also influences anti-trypanosomal activities. This is evident when comparing the effects of compounds 3a, 3b and 3c on parasite viability at a concentration of 20 µM. Compounds 3a and 3b bearing -NO 2 , and -Br, respectively, exhibited more than 75 % parasite growth inhibition, while compound 3c having -CF 3 substituent exhibited less than 5 % parasite growth inhibition. These results suggest that electron-withdrawing units promote activity while electron-donating unit leads to poor activity. The substitution pattern at position -3 also seems to greatly influence activity. Comparing the activity of compounds 3g (IC 50 , 24 µM) and 4g (IC 50 , 7 µM), both of which differ only in the substituent at position -3, suggests that an amide moiety at position -3 seems to enhance activity over ethyl ester.

Conclusion
We have synthesized a series of novel quinolones with varied substituents at position -1, -3 and -6 of the quinolone scaffold. While most compounds in this series showed no promising cytotoxicity potentials, compounds 4g emerged as potent anti-trypanosomal hit with IC 50 value of 7 µM. Although this series exhibited moderate to weak activities profiles, the comprehensive structure-activity relationship analyses of this series will undoubtedly serve as a resource for further optimization of the quinolone scaffold in search of new and potent anti-trypanosomal agents.

General Method
All the chemicals and solvents used were purchased from various chemical suppliers and were used without further purification. Melting points were determined using a Reichert hot stage microscope and are uncorrected. The progress of the reactions was monitored by thin layer chromatography (TLC) using Merck F254 silica gel plates supported on aluminium. The crude products were purified by a silica gel column chromatography using Merck Kieselgel 60 Å: 70-230 (0.068-0.2 mm) silica gel mesh. 1 H and 13 C NMR spectra were recorded on Bruker Biospin 300 MHz, or 400 MHz spectrometers, and the chemical shifts are given in d values referenced to solvents and are reported in parts per million (ppm). The high-resolution mass spectrometric data of final compounds was recorded on a Waters Synapt G2 quadrupole time-of-flight (QTOF) mass spectrometer operated with an electrospray ionization probe in the positive mode (University of Stellenbosch). The instrument was operated with an electrospray ionization probe in the positive mode. The starting quinolones 2 were synthesized from the synthetically accessible compounds 1 as previously described in literature. 34

General Method for the Preparation of N-alkylated Compounds 3a-h
A mixture of 2 (3.86 mmol, 1 g, 1 eq), K 2 CO 3 (5.0 mmol, 0.53 g), alkyl halide (5 eq.) in acetone (50 mL) was refluxed for 15 h. Upon reaction completion as indicated by TLC, the mixture was filtered, and the filtrate evaporated to dryness to obtain a crude N-alkylated product which was purified through silica gel column chromatography using CH 2 Cl 2 /MeOH (10:1) as the mobile phase. Compounds 3a-h were obtained in 40-70 % yield following this procedure.

General Method for the Preparation of Amides 4a-j
A mixture of 3 (1 g, 1 eq.), DBU (320 µL, 0.33 g, 2.1 mmol), an appropriate amine (5 eq.), and chloroform (15 mL) in a 100 mL round-bottom flask was stirred under reflux for 24-30 h. 34 Upon reaction completion as indicated by TLC, the mixture was evaporated to dryness and resultant crude subjected to silica gel column chromatography eluting with CH 2 Cl 2 /MeOH (10:1). Fractions containing the desired product were combined, evaporated to dryness and recrystallized from ethanol. Compounds 4a-j were obtained in 30-50 % yield following this procedure.