Discovery of a highly potent novel rifampicin analog by preparing a hybrid of the precursors of the antibiotic drugs rifampicin and clofazimine

Tuberculosis (TB) is an infectious disease caused by the bacillus Mycobacterium tuberculosis (Mtb). The present work reports the design and synthesis of a hybrid of the precursors of rifampicin and clofazimine, which led to the discovery of a novel Rifaphenazine (RPZ) molecule with potent anti-TB activity. In addition, the efficacy of RPZ was evaluated in-vitro using the reference strain Mtb H37Rv. Herein, 2,3 diamino phenazine, a precursor of an anti-TB drug clofazimine, was tethered to the rifampicin core. This 2,3 diamino phenazine did not have an inherent anti-TB activity even at a concentration of up to 2 µg/mL, while rifampicin did not exhibit any activity against Mtb at a concentration of 0.1 µg/mL. However, the synthesized novel Rifaphenzine (RPZ) inhibited 78% of the Mtb colonies at a drug concentration of 0.1 µg/mL, while 93% of the bacterial colonies were killed at 0.5 µg/mL of the drug. Furthermore, the Minimum Inhibitory Concentration (MIC) value for RPZ was 1 µg/mL. Time-kill studies revealed that all bacterial colonies were killed within a period of 24 h. The synthesized novel molecule was characterized using high-resolution mass spectroscopy and NMR spectroscopy. Cytotoxicity studies (IC50) were performed on human monocytic cell line THP-1, and the determined IC50 value was 96 µg/mL, which is non-cytotoxic.


Results
With the aim of developing better analogs of rifampicin, an analog of rifampicin named Rifaphenazine (RPZ) was synthesized in the present study by preparing a hybrid of the precursors of the approved drugs rifampicin (RIF) and clofazimine (CFZ). This analog was synthesized through a rapid single-step procedure illustrated in Fig. 1. Briefly, formyl rifampicin 3 was allowed to react with 2,3 diaminophenazine 4 in the presence of glacial acetic acid, producing Rifaphenazine (RPZ) 5 with a moderate yield of 53%. The structure of the synthesized analogs was characterized using high-resolution mass spectrometry, 1H NMR, 13C NMR, and 2D NMR techniques such as heteronuclear multiple bond correlation (HMBC) and heteronuclear single quantum correlation (HSQC) spectroscopy. The calculated molecular weight of the hybrid drug RPZ (C 50 H 53 N 5 O 12 ) was 915.3700, www.nature.com/scientificreports/ while the experimentally observed molecular weight in high-resolution mass spectroscopy (HRMS) was m/z 916.3951(M + 1)+. The antituberculosis activity of Rifaphenazine (RPZ) against the M. tuberculosis H37Rv strain was evaluated under in-vitro conditions using the broth microdilution method followed by spotting the drug-treated culture onto Middlebrook 7H11 agar supplemented with OADC (oleic acid, bovine serum albumin, dextrose, and catalase). The drug candidate RPZ at the drug concentration of 0.5 µg/mL exhibited 1.3log 10 reduction in the bacterial load compared to the control (without drug), while the positive control RIF exhibited 0.7log 10 difference only. At a lower drug concentration of 0.1 µg/mL, RPZ exhibited a difference of 0.7 log 10 compared to the control (without drug), while RIF at the same drug concentration did not exhibit any activity at all. Figure 2 depicts the graphical representation of drug vs. viable Mtb log 10 CFU/mL.
The drug efficacy, i.e., the concentration of the drug vs. the % inhibition, is presented in Fig. 2. A 78% Mtb inhibition was observed for RPZ at 0.1 µg/mL, while the positive control drug RIF did not exhibit any inhibition at the same concentration. Similarly, at a higher concentration of 0.5 µg/mL, RPZ exhibited Mtb inhibition of 93%, while RIF exhibited only 72% inhibition.

Minimum inhibitory concentration (MIC) study.
The results from the Minimum Inhibitory Concentration (MIC) experiments are presented in Table 1. The MIC study 17 was conducted against the Mtb H37Rv strain using the broth microdilution method. It was observed that the drug RPZ exhibited MIC at 0.5 µg/mL, while RIF exhibited MIC at 1 µg/mL. The reproducibility of these results was confirmed through another MIC study using the Mtb H37Rv strain that was passaged freshly and isolated from the guinea pig, and it was observed that both RPZ and RIF exhibited MIC at 1 µg/mL. Moreover, the clofazimine precursor Phenazine (4) did not exhibit any activity up to the concentration of 2 µg/mL.  www.nature.com/scientificreports/ Time-kill assay. Exposure of Mtb H37Rv to RPZ at a concentration of 0.5 μg/mL was observed to reduce the Mtb population by 3Log 10 within 24 h of drug exposure. Even after seven days of exposure, no emergence of the Mtb surviving population was observed, indicating the drug's Mtb-sterilizing property and stability. The time-dependent bactericidal property of RPZ is similar to that of the rifampicin drug. (Fig. 3).
Cytotoxicity assay. The cytotoxicity of RPZ was assessed ex-vivo using human monocytic cell line THP-1, and its IC 50 was estimated. RPZ presented an IC 50 value of 96 µg/mL, while RIF presented an IC 50 of 87 µg/mL (Table 1). With IC 50 > 90 μg/mL, RPZ could be classified as a non-cytotoxic compound.

Discussion
The present study reports the successful synthesis of the novel Rifaphenazine (RPZ) molecule and proposes it as a potential and safe anti-TB drug based on the in-vitro analysis. The RPZ molecule was able to arrest the growth of MTB even at a lower drug concentration. The rifampicin (RIF) drug generally acts by inhibiting the bacterial DNA-dependent RNA polymerase specifically, which is achieved through the formation of a stable complex of the drug with the ß-subunit of the RNA polymerase 18 . Since RPZ is structurally similar to RIF, it was expected that RPZ may also bind with RNA polymerase and inhibit the growth of Mtb. Besides, our in-silico molecular docking simulation study demonstrates that RPZ binds with the active site of Mtb RNA polymerase with a binding affinity of − 8.2 kcal/mol (Fig. 4).
Rifampicin is an important first-line anti-TB drug, while drugs such as clofazimine (CFZ) are used in secondline anti-TB treatment. However, CFZ has severe side effects, including hyper pigmented patches appearing on the skin 19 . The two chlorobenzene moieties attached to the phenazine core of CFZ might be the reason for these side effects. If chlorobenzene and dimethyl methane are removed from the structure of the CFZ drug using the  www.nature.com/scientificreports/ retrosynthetic approach, it produces 2,3-diamino phenazine at the main core of the CFZ drug. Therefore, the same 2,3-diamino phenazine was obtained from a commercial source, and its anti-TB activity was evaluated. However, the results revealed that the drug precursor (2,3-diamino phenazine) had not shown any anti-TB activity. Subsequently, it was hypothesized that, in order to recover the anti-TB activity that was lost due to the altered chemical properties of 2,3-diaminophenazine during the process of making the drug more effective and less toxic, it is necessary to form a hybrid of this drug precursor with another active pharmacophore, i.e., Rifampicin core. The reaction mechanism of the synthesis of Rifaphenazine 5 involved rapid condensation of the formyl group in rifampicin with an amino group in phenazine in the presence of glacial acetic acid as catalyst, followed by cyclization. Interestingly, the synthesized hybrid was observed to be more potent than the Rifampicin drug against the Mtb H37Rv strain. Furthermore, due to the aromatic nature of the attached phenazine ring, the synthesized analog (RPZ) was relatively more stable than RIF and could, therefore, be stored at ambient temperatures. The 1H NMR spectrum of RPZ revealed six aromatic protons of the phenazine ring that resonated between δ7.63 ppm and δ8.26 ppm, while the amide N-H proton appeared at δ11.79. The HMBC spectrum revealed that the N-H proton correlated with the naphthyl carbon atoms in the RIF core, while the 1H-13C of HSQC spectrum revealed that the six aromatic protons of the phenazine ring correlated with the carbon atom that appeared between δ106 and δ129.
In conclusion, these data suggested that Rifaphenazine (RPZ) is a promising pre-clinical candidate for the treatment of drug-susceptible tuberculosis (DS-TB). The results of the in-vitro analysis suggest that the efficacy of Mtb inhibition demonstrated by RPZ at a lower drug concentration has a prospect of shortening the TB treatment duration. The significance of the present work is that, even at a one-tenth concentration of rifampicin MIC, the drug candidate Rifaphenazine (RPZ) was able to kill 78% of Mtb, while rifampicin could kill zero percent Mtb at the same concentration. Furthermore, the cytotoxicity studies revealed that RPZ was not cytotoxic to the human monocytic cell line THP-1. Future research should endeavor to perform pre-clinical studies with different sets of bacterial strains and Mtb clinical strains to provide a strong lead to clinical studies.

Materials and methods
Chemistry. The synthesized compounds were characterized through 1H-NMR and 13C NMR using Bruker Avance 500 spectrometer and CDCl 3 as the solvent. High-resolution mass spectra (HRMS) were obtained by using Agilent Q-TOF-Mass Spectrometer in the ESI positive mode, and the values were expressed in m/z. The synthesized drug analog was purified through column chromatography using glass columns packed with silica gel (100-200 mesh). Thin Layer Chromatography (TLC) was performed using Merck aluminium-backed TLC Silica gel plate (60 F254). The chemicals and solvents used in the analyses were obtained from commercial sources and used as received.
General procedure for the synthesis of the RIF analog. Synthesis of Rifaphenazine 5. Formyl rifampicin 3 (1 equivalent) and 2,3 diamino phenazine 4 (1.2 equivalent) were added to a suspension of glacial acetic acid (100 µL) in MeOH (15 mL) at 45 ℃, followed by stirring for 2 h to allow the reaction to occur. After the completion of the reaction, the solvent was removed under vacuum and the solid product was extracted with Dichloromethane (DCM)/Water (3 × 10 mL). The collective DCM extracts were dried with sodium sulphate (Na 2 SO 4 ) and then concentrated to obtain the final product 5 in solid form. The produced compound was purified through column chromatography using the mixture of dichloromethane and methanol as solvent. TLC  13

The antituberculosis activity of RPZ against Mtb H37Rv. The Minimum Inhibitory Concentration
(MIC) study was performed against the M. tuberculosis H37Rv strains with different susceptibilities (standard H37Rv strain and the same strain passaged recently and isolated from guinea pig) using the broth microdilution method 17 , and the results are presented in Table 1. Briefly, the Mtb culture in Middlebrook 7H9 broth at the logphase of growth was adjusted to a McFarland standard turbidity of 0.5 and allowed to settle for fifteen minutes. Subsequently, this culture was added to the wells of a 96-well microtiter plate to reach the final drug concentrations ranging from 0.1 to 2.0 µg/mL. Among the wells, two wells represented drug-free (inoculum-only) controls. After 7 days of incubation, both drug-free and drug-treated wells were examined under an inverted phase-contrast microscope to observe the characteristic serpentine cord formation of Mtb. The lowest drug concentration that inhibited the microscopically-visible growth of Mtb after seven days was referred to as the Minimum Inhibitory Concentration (MIC) of the drug. In order to ascertain the bactericidal property of RPZ, the Mtb Colony-forming units (CFU) were enumerated by spotting a specified volume of serially-diluted aliquots drawn from the corresponding wells onto Middlebrook 7H11 agar supplemented with OADC [oleic acid, bovine serum albumin, dextrose, and catalase (Becton-Dickinson)]. The assay was performed in triplicate. The plates were incubated at 37 °C and the colonies were counted after 21 days of incubation.  21 . The proteins were prepared by removing all the water molecules and ligands using the Autodock Tools program version 1.5.6. The chemical structures of the drug were constructed using Chem3D. The energy minimization was conducted using the MM2 forcefield. A 120 × 120 × 120 and 3.75 Å grid box was set around the active sites for ligand interaction. The rigid grid box was prepared using Auto grid. The best docking conformation was obtained by using AutoDock with the Lamarckian Genetic Algorithm. The remaining docking parameters were set to default values. The binding pose presenting the best binding affinity was visualized using Pymol (Version 2.4.0, Schrödinger, LLC, https ://www.pymol .org). Figure 4 depicts the interaction between the inhibitor RPZ and a receptor Mtb RNAP, presenting the H-bonds formed by the inhibitor with the residues present in the active site of the receptor and the non-covalent interactions (− 8.2 kcal/mol) contributing to the Mtb inhibition.