Synthesis and study of new antitubercular compounds

Tuberculosis is regarded as one of the deadliest diseases in the world. It is a bacterial infection caused by some bacteria from the genus Mycobacterium, such as Mycobacterium tuberculosis. Some bacterial strains are multi-resistant or extensively-resistant against classical antibiotics. Consequently, there is a necessity to set up new strategies to prevent the spread of antibiotic resistant mycobacteria. Quinoline core is present in some antitubercular compounds. Indeed, bedaquiline (one of the last commercialized antitubercular compounds) is a diarylquinoline, which acts by inhibiting selectively the mycobacterial ATP synthase. This enzyme is required for the energetic metabolism of the cell and is a critical target to kill dormant strains. Mefloquine is a quinoline used as antimalarial compound but this molecule shows also antimycobacterial properties. Mefloquine can inhibit ATP synthase of Streptococcus pneumoniae, this inhibition may explain it antimycobacterial activity. The objectives of this work are designing, synthesizing, and evaluating new antitubercular compounds as quinoline derivatives (AQM). These molecules are expected to inhibit mycobacterial ATP synthase in order to fight latent forms of mycobacteria. The previous works of the research team have allowed to identify a lead compound which shows an MIC of 1 μM against M. tuberculosis MtbH37Rv strain. A pharmacomodulation of this lead compound will be shown here.


Introduction : generalities
Tuberculosis is caused mainly by Mycobacterium tuberculosis (Mtb).
Symptoms : cough, fever, weight loss, respiratory pain etc… The principal targets are lungs (71,5 % cases in 2004) but all organs can be affected.
10.4 million of people infected and 1.7 million of death in 2017  second cause of death among infectious diseases.
The co-infection by VIH and tuberculosis is a major issue: in 2016, 40% of death among seropositive people was attributable to tuberculosis. There is also a risk of interactions between antitubercular and antiretroviral drugs. The treatments are long (several months), expensive and induce some side effects

Introduction : resistance
There is a lack of safe therapeutical options to deal with MDR and XDR Mtb strains.

 Urgent need of new antitubercular compounds
Introduction : biological target -Bedaquiline, a diarylquinoline, target Mtb F1F0 ATP synthase.
-Mefloquine, an antimalarial quinoline, can inhibit Streptococcus pneumoniae ATP synthase ATP synthase is an interesting target to kill Mtb, especially in latent phase, because this enzyme is necessary for oxydative phosphorylation and so for the energetic metabolism of the cell. Inhibition of ATP synthase depends on stereochemistry: Bedaquiline (1S,2R) enantiomer is 630 fold less active than (1R,2S) enantiomer  an asymmetric synthesis is required  Quinoline core seems to be effective to inhibit ATP synthase

Introduction : objectives
The objectives of this project are designing, synthetising and evaluating new quinoline derivatives to treat replicating and latent form of Mtb.
Previous works of the team have allowed to identify a lead compound 1 which shows interesting activity against Mtb (MIC = 1 µM). Only R enantiomer is active.
A first generation of Amino-Quinoline Methanol (AQM) 2 has been synthesized by pharmacomodulation of the phenyl group in para position of this lead compound.

Introduction : pharmacomodulation (1)
Pharmacomodulation was designed thanks to a Craig Plot. Craig Plot is a tool used in physical organic chemistry that allows to quantify electronic properties (with Hammet substituent constant σ ) and lipophily properties (with partition coefficient π) of several chemical groups.
With this tool, we can choose some groups for this pharmacomulation with different physico-chemical properties in order to maximize the probability of finding the best properties for effective pharmacodynamic interactions with the target.  The desired AQMs 2 were obtained from a quinoline epoxyde 3 and the corresponding amine 4. The quinoline epoxyde was obtained from an hydroxy-quinoline 5 in four steps. This synthetic route is asymmetric, we have obtained a couple of enantiopur AQMs 5.
The hydroxy-quinoline 5 was reacted with POBr 3 to give the bromo-quinoline 6. A Suzuki coupling with vinyl trifluoroborate allowed to obtain the vinyl quinoline 7. A Sharpless dihydroxylation gave access to both enantiomers of diol 8. A final reaction of one-pot/3steps condensation gave both enantiomers of quinoline epoxyde 3 with retention of configuration.
This enantioselective synthetic route gave access to the S and R enantiomers of quinoline epoxyde 3 in four steps with a global yield of respectively 54% and 58%, and an enantiomeric excess of 93% and 96%.

Results and discussion : AQMs synthesis
13 Each enantiomer of quinoline epoxyde 3 was reacted with the amine 4 to form the desired AQM 2 with 26-89% yield. The reaction took place under microwawe heating, at 130 °C and 150 W during 30 min.
Twenty two molecules have been synthesized through this asymmetric synthetic route in 5 steps with a global yield from 15 to 50 % and an enantiomeric excess between 90 and 96%. Some AQMs have been tested in vitro, on atypical mycobacteria M. xenopi and M. avium.
These compounds are less active than Rifampicin and Bedaquiline. They show MIC between 8 and more than 32 mg/L on M. xenopi and are few active on M. avium (MIC >64 mg/L). 2c is the most active molecule, with the same activity than 1a on M. xenopi (8 mg/L).
Interestingly, we observe a difference of activity between R and S enantiomer : R is often more active.
On the Craig Plot graph, we can circle in red the chemical groups corresponding to AQMs with the best antimycobacterial activity for M. xenopi (8 to 16 mg/L) and circle in green the others.
Hydrophobics and electron-withdrawings groups (upper right quadrant of the graph) seems to give compounds with low activity.

By
contrary, electron-donor groups seems to allow to obtain quite good activity. Interestingly, the hydrophobicity of the group shows no or few difference concerning biological activity while the electron-donor effect is high enough.
Twenty two molecules have been synthesized with an enantioselective synthetic route in 5 steps with a global yield from 15 to 50 %. Some compounds (R= OH, Me and OMe) have shown encouraging activities (MIC = 8 to 16 mg/L on M. avium and M. xenopi). The difference of biological activity between the R and S enantiomers justify the necessity of an asymetric synthesis. We will continue to explore the corresponding zone of the Craig plot (electron/donor groups) in order to increase the biological activity.
When the best group is find in this pharmacomodulation, a second generation of AQM will be synthesized. This second generation will differ by the introduction of a phenyl group between amino and hydroxy function of the AQM core, in order to get closer to Bedaquiline's structure and to increase the lipophilicity, which will allow the AQMs to reach latent bacteria.
Others ideas of pharmacomodulation consist of modified aliphatic chain length between amino group and aromatic part, and the nature of the aromatic moiety.