Systematic Evaluation of Structure-Activity Relationships of the Riminophenazine Class and Discovery of a C2 Pyridylamino Series for the Treatment of Multidrug-Resistant Tuberculosis

Clofazimine, a member of the riminophenazine class of drugs, is the cornerstone agent for the treatment of leprosy. This agent is currently being studied in clinical trials for the treatment of multidrug-resistant tuberculosis to address the urgent need for new drugs that can overcome existing and emerging drug resistance. However, the use of clofazimine in tuberculosis treatment is hampered by its high lipophilicity and skin pigmentation side effects. To identify a new generation of riminophenazines that is less lipophilic and skin staining, while maintaining efficacy, we have performed a systematic structure-activity relationship (SAR) investigation by synthesizing a variety of analogs of clofazimine and evaluating their anti-tuberculosis activity. The study reveals that the central tricyclic phenazine system and the pendant aromatic rings are important for anti-tuberculosis activity. However, the phenyl groups attached to the C2 and N5 position of clofazimine can be replaced by a pyridyl group to provide analogs with improved physicochemical properties and pharmacokinetic characteristics. Replacement of the phenyl group attached to the C2 position by a pyridyl group has led to a promising new series of compounds with improved physicochemical properties, improved anti-tuberculosis potency, and reduced pigmentation potential.


Introduction
Tuberculosis (TB) infects about some 9.2 million and kills approximately 1.8 million people globally every year [1]. Despite the best treatment and control efforts using available vaccines and drugs, the prevalence of TB throughout the World remains high, due in part to the emergence of multidrug resistant Mycobacterium tuberculosis (MDR-TB) [2] and the global HIV epidemic. The treatment of MDR-TB and TB/HIV co-infection using the currently available drug regimens is extremely challenging [3]. New and more effective drugs are needed to address these challenges.
In an effort to identify new anti-tuberculosis agents that can be effective against MDR-TB, we decided to revisit the riminophenazine class of drugs. Clofazimine (CFZ, Figure 1), a member of the riminophenazine class, is the cornerstone agent for the treatment of leprosy. This agent was first synthesized by Barry's group [4] and has demonstrated potent in vitro activity against M. tuberculosis [5].
Recent studies indicate that clofazimine is equally potent against drug-susceptible and drug-resistant (including MDR) M. tuberculosis strains, which suggests that it operates via a novel mechanism of action. In addition, the frequency of resistance development for this compound is extremely low as compared with other anti-tuberculosis drugs [6]. Clofazimine is currently in clinical trials for the treatment of MDR-TB to address the urgent need for new drugs that can overcome existing and emerging drug resistance in tuberculosis. However, the widespread use of clofazimine in tuberculosis treatment is hampered by its extremely high lipophilicity and strong skin pigmentation side effects. Although a large number of analogues of clofazimine have been prepared, they are limited in scope. The modifications are limited to the substituents on two pendant phenyls and the imino group [7]. To identify a new generation of riminophenazines that is both less lipophilic and skin staining, while maintaining its efficacy against tuberculosis, a systematic structure-activity and structure-liability relationship (SLR) study is needed. Clofazimine is a lipophilic molecule as indicated by its high Clog P value of approximately 7.48 [8]. Its undesirable physicochemical and pharmacokinetic properties contribute to its side effects and limit its clinical use. In particular, its extremely long half-life and propensity for tissue accumulation and precipitation together with the dye property lead to unwelcome skin pigmentation. Herein we report our SAR and SLR studies on the riminophenazine class with a goal of identifying a new generation of compounds with improved physicochemical properties that are highly efficacious against both drug-susceptible and MDR-TB and at the same time incur reduced skin pigmentation potential.

Results and Discussion
To improve its physicochemical properties and circumvent the skin pigmentation problem of clofazimine, we first explored the possibility of decreasing the intrinsic colour of the molecules by reducing the conjugation system of the riminophenazine core structure. Thus, we systematically removed one of the phenyl rings (A, D or E) from the molecule while keeping rings B, C, and the imino moiety intact. The syntheses of these analogs are illustrated in Schemes 1 and 2, respectively. The synthesis of the A-ring deletion analogs (riminoquinoxalines) was started from 2,5-difluoronitrobenzene (Scheme 1). After replacement of the ortho-fluoro group with 4-chloroaniline, compound 1 was reduced to give diamine 2. Compound 2 was then treated with ethyl oxalyl chloride to form the B ring. After nitration of 3, the activated fluoro at the ortho position of the nitro adduct of 4 was replaced by 4-chloroaniline. Reduction of the nitro group in 5 produced compound 6, followed by reductive alkylation to provide compound 7. Reduction of the carbonyl groups in 7 with LiAlH 4 followed by an air auto-oxidization provided the target compound 8.  Both D-ring and E-ring deletion compounds 15a and 15b were synthesized starting from 1,5-difluoro-2,4-dinitrobenzene (DFDNB) (Scheme 2).
Thus compound 9a-b was reduced to diamine 10a-b which was then coupled with 1,5-difluoro-2,4-dinitrobenzene to give 11a-b. The second fluorine atom was replaced by 4-chloroaniline to afford 12a-b. Reduction of both nitro groups gave 13a-b which underwent spontaneous cyclization to afford riminophenazines 14a-b. Replacement of the imine with isopropylamine gave the target compounds 15a-b.
The A-ring deletion compound 8 exhibited considerably reduced in vitro activity against M. tuberculosis (MIC 90 = 18.89 μM, Table 1). When the appendant phenyl group D or E was replaced by a methyl group (compounds 15a and 15b), the antimycobacterial activity was also abolished (MIC 90 ≥ 42.45 μM and MIC 90 = 21.23 μM respectively, Table 1). Failing to identify analogues that retained antimycobacterial activity through dissecting their core structures, we turned our attention to the modification of the D-and E-rings of the molecule to improve its physicochemical properties and thereby change tissue distribution and tissue accumulation. It was anticipated that the Clog P of the molecule would decrease if one or both of the phenyl groups were replaced by a pyridyl group. Thus compounds 21a-b and 25 were synthesized following a modified synthetic procedure (Schemes 3 and 4).   In addition, compound 25 demonstrated excellent in vivo efficacy in a mouse tuberculosis model as measured by a significant drop in bacterial colony-forming units (CFU) in the lung as compared to a no-treatment control group [9]. The mean log CFU count in the lung dropped by about 5 units (3.83 ± 0.27 CFU/lung ) for the group treated with compound 25 as compared to the untreated group (8.53 ± 0.32 CFU/lung), and by about 3 units as compared to rifampin (RIF)-treated positive control groups (RIF 6.71 ± 0.13 CFU/lung, CFZ 6.33 ± 0.07 CFU/lung). Compound 25 also demonstrated a shorter half-life (t 1/2 ) as compared to clofazimine and reduced abdominal tissue pigmentation, which is strikingly different from the effect on those animals treated by clofazimine (data not shown). This SAR and SLR study illustrates that all five aromatic rings of clofazimine are important for maintaining potent anti-tuberculosis activity. The two pendant phenyl rings attached to the C2 and N5 positions, especially the C2 position, can be replaced by a pyridyl group, and such analogues have demonstrated potent in vitro anti-tubercular activity, excellent in vivo efficacy, and reduced skin pigmentation potential.

Reagents and Instrumentation
Unless otherwise indicated, all reagents and solvents were used as received from the suppliers. Reactions were monitored for completion by thin layer chromatography (TLC) using silica gel GF-254 plates with detection under UV (254 nm) light. 1 H-NMR spectra were recorded on a Varian 300 MHz or 400 MHz instrument in CDCl 3 , CD 3 OD, DMSO-d 6 or acetone-d 6 solutions. 13 C-NMR spectra were recorded on a Varian instrument at 100 MHz or 150 MHz with CDCl 3 or DMSO-d 6 as solvents. Chemical shifts are reported in parts per million (δ) downfield from tetramethylsilane (TMS). Coupling constants (J) are reported in Hz. High-resolution mass spectra were acquired from an Agilent 1100 series LC/MSD mass spectrometer. All MS experiments were performed using electrospray ionization (ESI-TOF + ) in positive ion mode. Column chromatography was carried out with silica gel (200~300 mesh). (1) 4-Chloroaniline (12.76 g, 100 mmol), 2,5-difluoronitrobenzene (8.0 g, 50 mmol), anhydrous K 2 CO 3 (3.5 g, 25 mmol) and anhydrous KF (2.9 g, 50 mmol) were mixed and heated at 170 °C for 14 h. After cooling to room temperature, water was added and the solid formed was filtered and washed with water. The crude product was purified via recrystallization from 95% ethanol to give 1 as an orange solid (11.8 g, 89%). 1

Synthesis of 1-(4-chlorophenyl)-6-fluoroquinoxaline-2,3(1H,4H)-dione (3)
Compound 1 above (6.45 g, 24 mmol) was suspended in anhydrous methanol (100 mL). The mixture was hydrogenated with 10% Pd/C (1.3 g) at room temperature at 1 atmosphere pressure of hydrogen for 8 h. After removal of catalyst, the solvent was concentrated under reduced pressure. The residue was dissolved in toluene (150 mL) and ethyl oxalyl chloride (10 mL, 89 mmol) was added. The mixture was refluxed for 1 h and then cooled to room temperature Approximately one half of the solvent was removed under reduced pressure and the resulting solid was filtered and recrystallized from methanol to give 3 as a white solid (4.97 g, 71%). 1 Compound 3 above (2.9 g, 10 mmol) was dissolved in concentrated sulfuric acid (30 mL) and cooled to −5 °C. KNO 3 (1.1 g, 8 mmol) was added into the solution portion-wise. After addition, the mixture was stirred at 0 °C for 1 h and then at room temperature for 1 h. The reaction mixture was slowly poured into ice water, and the solid formed was filtered and washed with water. The product was air-dried to give a light yellow solid 3.4 g (quantitative yield). 1 (7) Zinc powder (0.13 g) was added portionwise to a vigorously stirred mixture of 5 (0.08 g, 0.18 mmol) in glacial acetic acid (2 mL) and anhydrous methanol (5 mL). After the reaction was complete, as confirmed by TLC, the reaction mixture was filtered. The filtrate was concentrated under reduced pressure. Water was then added to the residue. After filtration, the yellow solid was dissolved in a mixture of glacial acetic acid (4 mL) and acetone (0.3 mL) in anhydrous methanol (2 mL). The resulting mixture was stirred at room temperature for 30 min., and then sodium borohydride (76 mg, 2.0 mmol) was added and stirred for 30 min. The reaction mixture was concentrated under reduced pressure and water was added to the residue. The resulting solid was collected by filtration, washed with water, and dried. The crude product was purified by column chromatography to give 7 as a solid (30 mg, 37%). 1  3.2.6. Synthesis of 1-(4-chlorophenyl)-6-(4-chloroanilino)-7-isopropylimino-1,7-dihydroquinoxaline (8) Lithium aluminum hydride (8 mg, 0.21 mmol) was added into a mixture of 7 (30 mg, 0.07 mmol) in THF (5 mL). The mixture was heated at 50 °C for 1 h under nitrogen atmosphere. After cooling to room temperature, 3 drops of ethyl acetate were added. The mixture was filtered and washed with ethyl acetate. The filtrate was washed with brine. The organic layer was separated and applied to preparative TLC. The TLC plate was heated in an oven (100 °C) for 30 min. before being developed with CHCl 3 /MeOH (20:1). The red band (R f : 0.34) was collected and washed with anhydrous methanol. The washing was concentrated, and the residue was further purified by column chromatography (HL20, eluted with MeOH) to give 6 mg of 8 as a red solid (21%). 1 1 g, 150 mmol), anhydrous KF (5.8 g, 100 mmol) and anhydrous K 2 CO 3 (13.8 g, 100 mmol) were heated at 160 °C for 10 h. The reaction mixture was cooled and water and ethyl acetate were added. The aqueous layer was extracted with ethyl acetate. The organic layer was combined and washed with 2 N HCl and dried over anhydrous Na 2 SO 4 . The solvent was concentrated under reduced pressure and the residue was recrystallized with anhydrous ethanol to give the title compound as red solid (20 g, 80%). 1

Synthesis of 2-(4-chloroanilino)-5-methyl-3-isopropylimino-3,5-dihydrophenazine (15a)
Zinc powder (1.96 g) was added portionwise to a suspension of 12a (0.48 g, 1.0 mmol) in glacial acetic acid (10 mL) and heated at 50 °C for 30 min. After filtration, the filtrate was stirred in contact with air overnight. The reaction mixture was concentrated under reduced pressure and adjusted to alkaline with ammonia, and then the solid formed was filtered and washed with water to give a black solid 0.34 g. The black solid was taken up in dioxane (5 mL) and isopropylamine (2.2 mL, 25.7 mmol) was added. The mixture was heated in a sealed bomb at 110 °C for 7 h. After being cooled to room temperature, water was added to the mixture and the solid formed was filtered and purified by column chromatography (eluted with P.E./ethyl acetate: 4/1 to 2/1 to 100% ethyl acetate) to give 0.13 g of the title compound (29%

Synthesis of 2-methylamino-5-(4-chlorophenyl)-3-imino-3,5-dihydrophenazine (14b)
Zinc powder (4.0 g, 61.5 mmol) was added portionwise to a suspension of 12b (0.8 g, 1.9 mmol) in glacial acetic acid (80 Ml) and stirred for 4 h. The reaction mixture was filtered and washed with glacial acetic acid. The filtrate was stirred in contact with air overnight. The solvent was concentrated and the residue was treated with water and adjusted to alkaline with ammonia. The solid formed was filtered, washed with water and dried. The crude solid was purified by neutral aluminum oxide (100-200 mesh) column chromatography eluted with ethyl acetate/methanol (10:1 to 5:1) to give the title compound as a dark red solid (0.28 g, 43%). M.p. 218-220 °C. 1

Synthesis of 2-methylamino-3-isopropylimino-5-(4-chlorophenyl)-3,5-dihydro-phenazine (15b)
A mixture of 14b (0.15 g, 0.45 mmol), isopropylamine (1 mL, 12.0 mmol) and dioxane (20 mL) was heated in a sealed tube at 120 °C for 10 h. After being cooled to room temperature, the reaction mixture was concentrated under reduced pressure. The residue was purified by neutral aluminum oxide Compound 17 (0.215g, 1.0 mmol) was suspended in anhydrous methanol (10 mL). Pd/C (10%, 40 mg) was added and the mixture was hydrogenated under atmospheric pressure until the reaction mixture turned colourless. The Pd/C was filtered off. The filtrate was concentrated in vacuo to give a light yellow oil. The oil was dissolved in THF (10 mL) and to this solution 16a (0.312 g 1.0 mmol) and diisopropylethylamine (0.17 mL, 1.0 mmol) were added. The mixture was refluxed for 20 h and then the reaction mixture was cooled to room temperature and concentrated in vacuo. The residue was added with water and filtered. The crude product was purified via chromatography (PE/EA: 2/1 to 1/1) to give the title compound as a yellow oil 0.243 g, yield 51%. 1

N1-(2,4-dinitro-5-(pyridin-3-ylamino)phenyl)-N2-(pyridin-3-yl)benzene-1,2-diamine (19b)
Compound 17 (0.215 g, 1.0 mmol) was suspended in anhydrous methanol (10 mL). Pd/C (10%, 40 mg) was added to the suspension and the mixture was hydrogenated under atmospheric pressure until the reaction mixture turned colourless. Then the Pd/C was filtered off. The filtrate was concentrated in vacuo to give a light yellow oil. The oil obtained was dissolved in THF (10 mL), and the solution was added 16b (0.278 g, 1.0 mmol) and diisopropylethylamine (0.17 mL, 1.0 mmol). The reaction mixture was refluxed for another 20 h. and then cooled to room temperature The mixture was concentrated in vacuo and the residues were added with distilled water and filtered. The crude product was purified via chromatography (PE/EA: 2/1 to 1/2) to produce the title compound as a yellow solid 0.3 g, yield 68%. 1

(E)-N-(4-chlorophenyl)-3-(isopropylimino)-5-(pyridin-3-yl)-3,5-dihydrophenazin-2-amine (21a)
Compound 19a (0.243 g, 0.5 mmol) was suspended in glacial acetic acid (10 mL). The suspension was added with zinc powder (0.66 g) portionwise in an ice bath. After addition, the mixture was stirred at room temperature for 30 min. and then heated at 50 °C for 30 min. After being cooled to room temperature, the reaction mixture was filtered and washed with glacial acetic acid. The filtrate was concentrated in vacuo. The residues were added with concentrated aqueous ammonia until the mixture became basic and then the solid was filtered out, washed with water and dried to give black solid. The black solid was dissolved in anhydrous methanol and the solution was added with solution of ammonia in methanol to adjust the solution to basic. The mixture was stirred in contact with air overnight and then concentrated in vacuo. The residues were mixed with dioxane (5 mL) and isopropylamine (2 mL, 24 mmol). The mixture was heated at 110 °C in a sealed tube for 10 h. After being cooled to room temperature, the reaction mixture was added with water and filtered to give crude product. The crude product was purified via chromatography (PE/EA: 2/1 to EA) to produce 21 mg of the title compound as a red solid, yield 9%. Mp: 237-239 °C. 1

(E)-3-(isopropylimino)-N,5-di(pyridin-3-yl)-3,5-dihydrophenazin-2-amine (21b)
Compound 19b (0.3 g, 0.7 mmol) was suspended in glacial acetic acid (5 mL). Zinc powder (1.33 g) was added portionwise to the suspension in an ice bath. After addition, the mixture was heated at 50 °C for 1 h. After being cooled to room temperature, the reaction mixture was filtered and washed with glacial acetic acid. The filtrate was concentrated in vacuo. The residues were added with concentrated aqueous ammonia until the mixture became basic and then the solid was filtered out, washed with water and dried to give a dark brown solid. The solid was dissolved in anhydrous methanol and the solution was added with ammonia in methanol to adjust the solution to basic. The mixture was stirred in contact with air for 7 h. and then concentrated in vacuo. The residues were mixed with dioxane (5 mL) and isopropylamine (0.036 mol, 3 mL). The mixture was heated at 110 °C in a sealed bomb for 10 h. After being cooled to room temperature, the reaction mixture was added with water and filtered to give crude product. The crude product was purified via chromatography (PE/EA: 2/1 to EA) to produce 40 mg of the title compound as a red solid, yield 15%. Mp: 174-175 °C.

Synthesis of 2-(3-pyridylamino)-5-(4-chlorophenyl)-3-imino-3,5-dihydrophenazine (24)
Zinc powder (71 g, 1.1 mol) was added portionwise to a mixture of compound 22 (28.6 g, 60 mmol) in glacial acetic acid (150 mL) cooled in an ice water bath. The mixture was stirred until the colour turned to light green, and then filtered, washed with glacial acetic acid and anhydrous methanol. The filtrate was concentrated and the residue was treated with water and adjusted to alkaline with ammonia. The solid formed was filtered, washed with water and then dissolved in anhydrous methanol. The methanol solution was stirred in contact with air overnight. The solid formed was filtered to produce 23.1 g of crude 24 which was used in the next step without further purification.

Biological Evaluation
In vitro and in vivo anti-tuberculosis activities were determined using our routine methods [9].

Conclusions
In summary, we have performed a systematic SAR and SLR investigation by synthesizing and evaluating a variety of analogues of clofazimine with modifications at various positions. The study revealed that the central tricyclic ring system is the pharmacophore of the molecule and is important for anti-tuberculosis activity. Attempts to simplify this pharmacophore led to compounds with reduced antimycobacterial activity. The two phenyl rings appended to the C2 and N5 positions can be replaced by a pyridyl ring. Replacement of the phenyl group attached to the C2 position by a pyridyl group leads to compounds with not only improved in vitro and in vivo anti-tuberculosis activity, but also favourable pharmacokinetic profiles with reduced skin pigmentation potential.