Pharmacokinetic and In Vivo Efficacy Studies of the Mycobactin Biosynthesis Inhibitor Salicyl-AMS in Mice

Abstract

Mycobactin biosynthesis in Mycobacterium tuberculosis facilitates iron acquisition, which is required for growth and virulence. The mycobactin biosynthesis inhibitor salicyl-AMS [5′-O-(N-salicylsulfamoyl)adenosine] inhibits M. tuberculosis growth in vitro under iron-limited conditions. Here, we conducted a single-dose pharmacokinetic study and a monotherapy study of salicyl-AMS with mice. Intraperitoneal injection yielded much better pharmacokinetic parameter values than oral administration did. Monotherapy of salicyl-AMS at 5.6 or 16.7 mg/kg significantly inhibited M. tuberculosis growth in the mouse lung, providing the first in vivo proof of concept for this novel antibacterial strategy.

TEXT

Siderophore production and iron acquisition by Mycobacterium tuberculosis are important for growth under iron-limited conditions and within human macrophages (1). They are also essential for the virulence of M. tuberculosis during infection (2). Importantly, gene expression profiling indicates that free iron is limited in human lungs (3). Hence, M. tuberculosis must compete for iron that is retained by high-affinity host proteins such as transferrin and lactoferrin. The siderophore-dependent iron acquisition system of M. tuberculosis consists of two major siderophore variants, cell wall-associated mycobactin and secreted soluble carboxymycobactin, both of which bind iron with high affinity (4, 5). It is believed that carboxymycobactin binds iron in the extracellular aqueous environment and transfers the iron to the cell wall-associated mycobactin that then transports the iron across the cell wall into the cytoplasmic space (5, 6). Biosynthesis of mycobactins is carried out by multiple enzymes, most of which are encoded by a cluster of genes designated mtbA to mtbJ (7). Systematic mutational studies of nine of the mbt genes (mbtA to mtbH and mbtT) in the Mycobacterium smegmatis model indicate that eight of the genes (all but mbtH) are required for mycobactin production (8). MbtA, a bifunctional enzyme encoded by the mbtA gene (Rv2384 in M. tuberculosis), catalyzes the initiation of mycobactin biosynthesis through a two-step reaction involving ATP-dependent adenylation of salicylate to generate a salicyl-AMP intermediate, followed by salicylation of an aroyl carrier protein domain of the nonribosomal peptide synthetase MbtB (7, 9). The importance of the mycobacterial siderophore biosynthesis pathway has led to the development of a designed small-molecule inhibitor, salicyl-AMS [5′-O-(N-salicylsulfamoyl)adenosine] (see Fig. S1 in the supplemental material), that specifically targets MbtA by mimicking the salicyl-AMP intermediate (9–11). In vitro studies have confirmed that salicyl-AMS inhibits MbtA activity and also inhibits M. tuberculosis growth in iron-depleted media (9, 12), with the MIC for M. tuberculosis H37Rv found to be 0.5 μg/ml in a study with an alamarBlue assay (13). In this study, we have characterized the pharmacokinetics of salicyl-AMS and its therapeutic efficacy in a murine infection model, leading to the first in vivo proof of concept for the use of siderophore biosynthesis inhibitors as novel antibacterials.

Compound bioavailability is essential for in vivo efficacy. In order to investigate the bioavailability of salicyl-AMS, we conducted single-dose pharmacokinetic studies with mice via two routes, oral gavage and intraperitoneal (i.p.) injection. Salicyl-AMS sodium salt was synthesized on a large scale via an optimized route (see the supplemental material for the synthesis procedure used) and dissolved in 1× phosphate-buffered saline (PBS) at a concentration of 5 or 20 mg/ml. Twenty-gram female BALB/c mice (Charles River Laboratories, Wilmington, MA) were given a single dose of either 50 or 200 mg/kg, either orally or by i.p. injection, in a volume of 0.2 ml. At 7, 15, 30, 60, and 120 min after compound administration, animals were euthanized and cardiac blood (∼0.7 ml) and lungs were collected. Plasma was separated by centrifugation at 12,000 rpm (benchtop microcentrifuge) for 20 min at 4°C. Mouse lungs were homogenized in 0.5 ml of liquid chromatography (LC)-mass spectrometry (MS) grade water. Salicyl-AMS concentrations in plasma and lung homogenate supernatants were analyzed by LC-tandem MS (LC-MS/MS; AB SCIEX QTRAP 5500 system) with, as an internal standard, stable-isotope-labeled salicyl-AMS synthesized from uniformly ¹³C₁₀- and ¹⁵N₅-labeled adenosine (Cambridge Isotope Labs; see the supplemental material for the synthesis procedure used) with detection of mass transitions 467.0/135.8 and 482.1/146.1. Data analysis was done with GraphPad Prism 4. Assays detected free salicyl-AMS in the plasma and lungs after both oral and i.p. administrations, indicating that this compound is bioavailable in mice. The lung-to-plasma exposure ratio (lung area under the concentration-time curve [AUC_lung]/plasma AUC [AUC_plasma]) was 0.52 at 50 mg/kg and 0.66 at 200 mg/kg. Similar kinetic patterns were observed after single-dose administration of salicyl-AMS between the two sample types, plasma and lung, and more interestingly, between the two administration routes, oral gavage and i.p. injection (see Fig. S2 and S3 in the supplemental material; Table 1). Although salicyl-AMS is bioavailable after oral administration, the concentrations achieved in plasma and lung samples after oral gavage are marginal compared to those obtained after i.p. injection, as indicated by the dose-normalized relative bioavailability (Table 1). In general, there are two in vivo working patterns for antimycobacterial agents, time over MIC dependent and maximum drug concentration in serum (C_max) dependent (14). The findings in this study are important because these data, i.e., C_max = 1.2 μg/ml and AUC = 58.6 μg · min/ml (Table 1), clearly suggest that oral administration of this inhibitor is less likely to achieve an effective systemic level, regardless of the in vivo mode of action of salicyl-AMS. This also highlights the need to optimize the bioavailability of this lead compound after oral administration in future work, because oral administration is preferable for tuberculosis drugs.

Table 1.

Single-dose pharmacokinetics of salicyl-AMS in mouse

Sample and dose (mg/kg)	Cmax (μg/ml)a		Tmax (min)a		AUC (μg · min/ml)a		Frel (%)b
	Oral	I.p.	Oral	I.p.	Oral	I.p.
Plasma
50	0.3 ± 0.05	64.6 ± 1.76	11.0 ± 5.66	7.0 ± 0.00	15.9 ± 0.59	1,280.0 ± 261.6	1.24
200	1.2 ± 0.96	329.7 ± 109.0	7.0 ± 0.00	7.0 ± 0.00	58.6 ± 12.60	9,185.0 ± 1,105.0	0.6
Lungc
50	0.4 ± 0.11	29.2 ± 7.29	7 ± 0.00	7 ± 0.00	12.3 ± 0.72	671.7 ± 163.8	1.83
200	1.9 ± 1.80	356.2 ± 72.10	7 ± 0.00	7 ± 0.00	56.4 ± 10.67	6,020 ± 369.1	0.94

^aData are shown as means ± standard deviations of three mice.

^bRelative bioavailability was calculated as follows, where D is the dose: F_rel = 100 × [(AUC_oral × D_i.p.)/(AUC_i.p. × D_oral)].

^cData for lungs were normalized to μg/g for C_max and μg · min/g for AUC.

Next, we examined the in vivo efficacy of salicyl-AMS in a mouse aerosol infection model. Thirty-five 6-week-old female BALB/c mice (Charles River Laboratories, Wilmington, MA) were aerosol infected with 10 ml of a late-log-phase M. tuberculosis H37Rv culture by an inhalation exposure system (Glas-Col Inc.). Five mice were sacrificed at day 1 for implantation determination, which was found to be 3.31 log₁₀ per mouse lung. Mice were randomly grouped to low-dose (5.6 mg/kg), high-dose (16.7 mg/kg), and untreated groups. From day 1 onward, mice were treated by i.p. injection of 0.2 ml of a PBS solution containing salicyl-AMS sodium salt with the corresponding dose, daily (5 days per week). At 2 and 4 weeks after treatment initiation, five mice in each group were sacrificed and their lungs were removed. All of the animal procedures used in this study were approved by the Institutional Animal Care and Use Committee of the Johns Hopkins University School of Medicine. The lungs were homogenized by bead beating, diluted with 1× PBS, and plated on Middlebrook 7H11 selective agar plates (Becton Dickinson & Co., Franklin Lakes, NJ). CFU per lung were enumerated after incubation for 4 weeks. Data are shown as means and standard deviations (Table 2). One-way analysis of variance (ANOVA) was carried out, and Tukey's multiple-comparison test was conducted. These data demonstrated that the mycobactin biosynthesis inhibitor salicyl-AMS significantly inhibited the growth of M. tuberculosis in mouse lungs, as shown by 0.87- and 1.10-log₁₀ lower CFU burdens in low-dose- and high-dose-treated mice, respectively, compared with those in untreated controls after 2 weeks of treatment (P < 0.01, Table 2). The CFU data also showed a dose dependency at this time point. After 4 weeks of treatment, although the lung CFU burden in the low-dose group was still lower than that in the untreated group, the difference was not statistically significant.

Table 2.

In vivo efficacy as represented by lung CFU burden

Time point	Mean lung CFU (log10) ± SDa
	Untreated control	Salicyl-AMS at:
		5.6 mg/kg	16.7 mg/kg
Wk 0	3.31 ± 0.022	3.31 ± 0.022	3.31 ± 0.022
Wk 2	7.92 ± 0.105	7.05b ± 0.084	6.82b ± 0.316
Wk 4	8.40 ± 0.669	7.72 ± 0.049	NAc

^aMean and SD from four mice.

^bP < 0.01 compared with untreated group as analyzed with one-way ANOVA and Tukey's multiple comparison test.

^cNA, not available.

Our study also provides insights into the toxicity profile of salicyl-AMS. Previous studies have shown that salicyl-AMS is nontoxic to the murine leukemia cell line P388 at concentrations of >200 μM (10, 11). We have also confirmed this lack of toxicity for Chinese hamster ovary cells at concentrations of up to 500 μM (data not shown). Notably, however, all of the mice treated with the higher dose (16.7 mg/kg) of salicyl-AMS died before the week 4 time point (Table 2). This drug toxicity could also contribute to the decreased efficacy observed at week 4 in the low-dose group. Since MbtA has no mammalian homologues, this toxicity is likely to be the result of off-target effects and/or brought about by a drug metabolite(s) generated in vivo. Further lead optimization of salicyl-AMS through medicinal chemistry may address the animal toxicity. The less-than-ideal in vivo efficacy of salicyl-AMS could also be attributable to its rapid clearance, as the calculated lung half-lives were only 13.3 and 19.3 min for 50 and 200 mg/kg, respectively (based on 30-, 60-, and 120-min data points). Despite these rapid clearance pharmacokinetic parameter values, salicyl-AMS still demonstrated in vivo efficacy, suggesting that MbtA inhibition is a promising antimycobacterial strategy.

In conclusion, we have characterized the pharmacokinetics and in vivo efficacy of salicyl-AMS in a murine model of tuberculosis. Inhibition of the mycobactin biosynthesis pathway by targeting of MbtA with salicyl-AMS slowed the growth of M. tuberculosis in mouse lungs. Although mouse infection models do not exactly reproduce human tuberculosis pathologies such as granuloma structures and cavities, they are the most convenient and cost-effective animal models with which to study the pharmacokinetics and efficacy of new tuberculosis therapies, and optimized leads can be advanced to nonhuman primate models for further assessment in the future. Lead optimization efforts to achieve lower drug toxicity, a longer half-life, higher efficacy, and better bioavailability after oral administration are ongoing.