Discovery of (3-Benzyl-5-hydroxyphenyl)carbamates as New Antitubercular Agents with Potent In Vitro and In Vivo Efficacy

A series of 3-amino-5-benzylphenol derivatives were designed and synthesized. Among them, (3-benzyl-5-hydroxyphenyl)carbamates were found to exert good inhibitory activity against M. tuberculosis H37Ra, H37Rv and clinically isolated multidrug-resistant M. tuberculosis strains (MIC = 0.625–6.25 μg/mL). The privileged compounds 3i and 3l showed moderate cytotoxicity against cell line A549. Compound 3l also exhibited potent in vivo inhibitory activity on a mouse infection model via the oral administration. The results demonstrated 3-hydroxyphenylcarbamates as a class of new antitubercular agents with good potential.


Introduction
Tuberculosis (TB) is a dangerous disease caused by M. tuberculosis (Mtb). Today it still poses the serious threat to the human health. According to the report of World Health Organization (WHO), in 2018 there were more than 10.0 million people newly infected with Mtb worldwide [1] and the number of deaths among TB patients was greater than that of HIV patients. Efficient antitubercular agents were developed to fight TB. Typically, the treatment includes a combination of 2-4 antitubercular agents for 6-24 months. The long therapeutic period led to a poor compliance, which readily induced the drug resistance of Mtb. Multidrug-resistant tuberculosis (MDR-TB) and extensively drug-resistant TB (XDR-TB) were reported to spread over the world and brought new challenges for TB treatment [2][3][4]. In 2018, 560,000 people were infected with rifampin-resistant or MDR-TB. Among them, 230,000 patients died from the disease. In addition, over 30% patients infected with XDR-TB were not curable [5]. In the past 30 years, only bedaquiline [6] and delamanid [7] were marketed for the treatment of MDR-TB. Several candidates including pretomanid [8] and SQ109 [9], are still in the clinical study. Bedaquiline is toxic, mainly for causing irregular heart rhythm [10]. Furthermore, Mtb strains resistant to these new drugs have also been found [11,12]. The situation poses an urgent need for antitubercular agents with new structures and new mechanisms of action [13].
In the previous study, we found that m-amidophenol derivatives YZ-6 and YZ-7 showed potent in vitro inhibitory activity against Mtb H37Ra, H37Rv and clinically isolated MDR-Mtb strains (MIC = 0.39-3.125 µg/mL) (Scheme 1) [14,15]. However, the two compounds did not exert in vivo = 0.39-3.125 μg/mL) (Scheme 1) [14,15]. However, the two compounds did not exert in vivo efficacy on a mouse infection model. Recently, we designed and synthesized a series of new m-amidophenol derivatives with 5-heteroatomic substitutents (OR, SR, NRR′) [16]. One privileged compound YQ-14 showed potent in vitro antitubercular activity and weak in vivo efficacy (Scheme 1). Because the clearance rates of YZ-6 and YQ-14 in mouse liver microsome were high (T1/2 = 6.60 and 5.43 min, respectively), further structural modifications to improve the metabolic stability and in vivo potency are desirable. We designed and synthesized new derivatives of YZ-6 and YZ-7 via the replacement of 1-amido group with sulfamide, urea and carbamate moieties. Such replacements were expected to provide a higher metabolic stability towards enzymatic hydrolysis. In addition, binding with the target may be enhanced considering the introduction of more oxygen and nitrogen atoms. Among the new derivatives, 3-hydroxylphenylcarbamates were found to exhibit potent in vitro and in vivo antitubercular efficacy. The results are reported in this paper. Scheme 1. Design of new 3-amino-5-benzylphenol derivatives.

Antitubercular Activity Against Mtb H37Ra
Antitubercular activity of 1a-1b, 2a-2b and 3a-3p was assessed against avirulent autoluminescent Mtb H37Ra and the results are listed in Table 1 and Table 2. The growth of the bacteria was monitored by the bioluminescence intensity [17]. The minimal inhibitory concentrations (MIC) were determined as the lowest concentration at which the RLU was 90% lower than the drugfree control. Isoniazid (INH) was used as the positive control.

Antitubercular Activity against Mtb H37Ra
Antitubercular activity of 1a-1b, 2a-2b and 3a-3p was assessed against avirulent autoluminescent Mtb H37Ra and the results are listed in Tables 1 and 2. The growth of the bacteria was monitored by the bioluminescence intensity [17]. The minimal inhibitory concentrations (MIC) were determined as the lowest concentration at which the RLU was 90% lower than the drug-free control. Isoniazid (INH) was used as the positive control.

In Vitro Inhibitory Activity Against Mtb H37Rv and MDR-Mtb Strains
The compounds 3i and 3l were selected for the further evaluation of inhibitory activity against Mtb H37Rv and clinically isolated multidrug-resistant Mtb strains (MDR-Mtb). The results are summarized in Table 4. Good inhibitory activities (MIC = 2-8 µ g/mL) were observed against Mtb H37Rv and MDR-Mtb strains K12, K16, K18, V4. In order to identify the key pharmacophores, we kept 4,4-dimethyl-cyclo-hexyl carbamate scaffold and replaced 1-hydroxyl with other functional groups. The inhibitory activity of compounds 3q and 3r against Mtb H37Ra are examined and the results are summarized in Table 3. The 1carboxylic analog 3q is inactive (MIC = 100 g/mL), however the methoxy analog 3r showed good inhibitory activity (MIC = 2.5 g/mL).

In Vitro Inhibitory Activity Against Mtb H37Rv and MDR-Mtb Strains
The compounds 3i and 3l were selected for the further evaluation of inhibitory activity against Mtb H37Rv and clinically isolated multidrug-resistant Mtb strains (MDR-Mtb). The results are summarized in Table 4. Good inhibitory activities (MIC = 2-8 µ g/mL) were observed against Mtb H37Rv and MDR-Mtb strains K12, K16, K18, V4.

In Vitro Inhibitory Activity against Mtb H37Rv and MDR-Mtb Strains
The compounds 3i and 3l were selected for the further evaluation of inhibitory activity against Mtb H 37 Rv and clinically isolated multidrug-resistant Mtb strains (MDR-Mtb). The results are summarized in Table 4. Good inhibitory activities (MIC = 2-8 µg/mL) were observed against Mtb H 37 Rv and MDR-Mtb strains K12, K16, K18, V4.

Evaluation of Cytotoxicity and Metabolic Stability
The cytotoxicity of 3i and 3l against human lung cancer cell line A549 was evaluated via MTT method and the results are summarized in Table 5. Moderate cytotoxicities were observed for compounds 3i and 3l (IC 50 = 15.8 and 22.6 µg/mL). The selection indices (SI = IC 50 /MIC) are 12.6 and 36.2 respectively for 3i and 3l. The result demonstrated compound 3l as a more suitable candidate for further evaluation. The metabolic stability of 3l in mouse liver microsome was examined. A quick clearance was observed (T 1/2 = 1.76 min, CL = 1440 mL/min/gprot). The result indicated that the replacement of 1-amido group in the lead compounds YZ-6 and YZ-7 with carbamate moiety cannot improve the metabolic stability. The reason remains to be investigated.

In Vivo Inhibitory Activity against Mtb H37Ra
The in vivo efficacy of compound 3l was evaluated in a mouse model infected with autoluminescent Mtb H37Ra. The compound 3l was administered orally (100 mg/kg/day). The untreated control group was administered with sodium carboxymethyl cellulose (CMC-Na) solution. Rifampicin was used as the positive control. After five days treatment, the burdens of the bacteria in lungs and spleens of the mice were determined by the bioluminescence intensity detection. The results are summarized in Figure 1. A significant reduction (1.3 log) of relative light unit (RLU) in the lung was achieved compared with the untreated control group. A small reduction of RLU (0.4 log) was observed in the spleen.
The metabolic stability of 3l in mouse liver microsome was examined. A quick clearance was observed (T1/2 = 1.76 min, CL = 1440 mL/min/gprot). The result indicated that the replacement of 1amido group in the lead compounds YZ-6 and YZ-7 with carbamate moiety cannot improve the metabolic stability. The reason remains to be investigated.

In Vivo Inhibitory Activity against Mtb H37Ra
The in vivo efficacy of compound 3l was evaluated in a mouse model infected with autoluminescent Mtb H37Ra. The compound 3l was administered orally (100 mg/kg/day). The untreated control group was administered with sodium carboxymethyl cellulose (CMC-Na) solution. Rifampicin was used as the positive control. After five days treatment, the burdens of the bacteria in lungs and spleens of the mice were determined by the bioluminescence intensity detection. The results are summarized in Figure 1. A significant reduction (1.3 log) of relative light unit (RLU) in the lung was achieved compared with the untreated control group. A small reduction of RLU (0.4 log) was observed in the spleen.

Chemistry
The NMR spectra of 1 H and 13 C were recorded on Bruker AVANCE 400 or 500 spectrometer (Karlsruhe, Germany). Chemical shifts of protons are reported in parts per million downfield from tetramethylsilane. Peaks are labeled as singlet (s), broad singlet (br), doublet (d), triplet (t), double doublet (dd), doublet of triplets (dt), multiplet (m). The high-resolution mass spectra were analyzed on a SHIMADZU LCMS-IT-TOF mass spectrometer (Tokyo, Japan). Melting points were determined in open capillary tubes on a MPA100 Optimelt automated melting point system. All chemicals were purchased from Sigma-Aldrich and Alfa Aesar chemical companies and were used without further purification. Compounds 4-8 were prepared according to the reported procedures [14].

Evaluation of Cytotoxicity
The cytotoxicities of compounds 3i and 3l were assayed against human lung cancer cell line A549 at concentrations from 100 to 6.25 µg/mL. Approximately 4 × 10 4 cells, suspended in culture medium, were seeded in the 96-well plate and then allowed to recover for 24 h at 37 • C under 5% CO 2 . The solution of tested compound was added to the plate and the final concentrations were 100 mg/mL, 50 mg/mL, 25 mg/mL, 12.5 mg/mL, and 6.25 mg/mL, respectively. Each experiment was repeated three times. After being incubated for 72 h, Fresh MTT was added to each well at a final concentration of 0.5 mg/mL and incubated with cells at 37 • C for 4 h. The formazan crystals were dissolved in 150 µL DMSO per each well. The absorbance at 490 nm was measured by microplate microscopy, and the survival rate was calculated by the formula (OD in the experimental group − blank group)/(OD in the cell group − blank group) × 100%. The cytotoxicity is shown as IC 50 value, which was calculated by GraphPad Prism Software version 5 [18].

Determination of the Clearance Rate in Mouse Liver Microsome
The mouse liver microsome was purchased from Xenotech company. The liver microsome was incubated in a 96-well plate. The final incubation volume was set at 45 µL, contained NADPH (2 mM), microsomes (0.5 mg/mL), the compound 3l (1 µM), MgCl 2 (5.0 mM), DMSO (0.01%) and BSA (0.005%). The plate was incubated at 37 • C for ten minutes, and then was added NADPH to trigger the reaction. The incubation times were 0, 5, 15, 30, 45 and 60 min. Ice-cold acetonitrile was added to terminate the reaction. The mixture was centrifuged at 4 • C for 15 min at 4000 rpm. The concentration of the compound 3l in supernatant was determined by LC-MS/MS. The incubation time was plotted with the logarithm of the drug residual rate in the incubation system. k was calculated by linear regression, and the clearance rate was calculated based the formula Cl int = 1000×slope P [19].

In Vivo Inhibitory Activity against Mtb H37a
The experimental program involving animal feeding and use was submitted to, and approved by, the Animal Welfare Department of Guangzhou Institute of Biomedicine and Health, Chinese Academy of Sciences.
Mtb H37Ra isolated on plates were homogenized with sterile glass beads in a 250 mL flask containing 50 mL 7H9 with tween80. When RLU reached 2 million/mL, the broth culture was used to infect 4-6-week-old male BALB/c mice by tail vein injection. The day after infection (day 0), RLU counts were determined. The mice were first anesthetized by isoflurane inhalation and the RLU count was determined by laying the breast of the mouse on the detection hole of the luminometer and measuring light production twice for 3 s. Mice with similar RLU readings were randomly allocated to treatment groups (4 mice/group) and individually marked. The treatment groups received: 0.5% CMC-Na alone as negative control; RIF (10 mg/kg) as positive control; 3l (100 mg/kg). The treatment was administered once daily by oral gavage for six days. On the final day of treatment, animals were