Synthesis and antibacterial activity of novel benzodioxin-containing oxazolidinones against M. abscessus

Abstract

Mycobacterium abscessus infections represent a significant treatment challenge, owing to the lack of effective therapies, and the intensive multidrug regimens currently recommended for this indication. Oxazolidinone antibiotics such as linezolid have been used to treat this complex disease. Utilizing a strategic structural extension strategy, five novel oxazolidinone derivatives were synthesized by the introduction of oxime, nitrile, amide, amidoxime, and oxime ester functional groups to a previously reported benzodioxin-scaffold. The in vitro activities of these derivates against M. abscessus were assessed, with the amidoxime derivative exhibiting activity superior to linezolid.

Graphical Abstract

Oxazolidinone antibiotics are a class of synthetic antibacterial compounds with clinical success in treating antimicrobial resistant bacteria ^1,2. The antibacterial activity of molecules containing an oxazolidinone moiety was first identified by researchers at DuPont de Nemours & Co in the late 1970s ^1,3. As interest in novel classes of antibiotics surged with the emergence of antimicrobial resistance in clinical settings, the first-in-class compound, linezolid (1) (Figure 1), was synthesized at the Upjohn Company in 1993, and was approved by the FDA in 2000 ^4–6.

Figure 1.

Current FDA-approved oxazolidinones and previously synthesized oxazolidinone T145 (3)

The antibacterial mechanism of action of oxazolidinones operates by inhibiting protein synthesis via binding within the peptidyl transferase center (PTC) of the bacterial ribosome ^6–8. Within the binding pocket of the PTC, oxazolidinones have been observed forming a number of conserved binding interactions with the structural ribosomal RNA (rRNA) ^9–13. In addition, it has recently been elucidated that their ability to bind is dependent on the identity of the amino acid occupying the penultimate position of the peptide chain being synthesized ^9,10. The transient nature of this binding environment, in conjunction with the difficulties associated with computationally modeling the target ribosome, has thus far rendered docking studies with oxazolidinones ineffective, despite their prevalence in the literature ¹⁴. In structures of linezolid bound to the PTC in the 50S subunit in numerous bacteria, there remains space for new interactions with rRNA within the binding pocket, meaning increased binding affinity can be achieved through strategic structural extension of the linezolid framework ^2,15. Other oxazolidinones that use a structural extension strategy at what is commonly referred to as the C-ring (Figure 2), such as the FDA-approved tedizolid (2) (Figure 1), or the candidate drugs radezolid and contezolid, have demonstrated improved activity against a number of pathogenic bacteria, including Mycobacterium abscessus ^2,10,16.

Figure 2.

Oxazolidinone antibiotic ring naming conventions, as applied on the left to Tedizolid (2) and on the right to compounds reported herein

Structural similarities between bacterial and human mitochondrial ribosomes lead to off-target binding of oxazolidinones to the human mitochondrial ribosome, resulting in significant dose-dependent toxicity ^2,10. As such, oxazolidinones are primarily utilized in clinical settings as a last resort for severe bacterial infections ¹⁷. It has also been observed that the introduction of additional functional groups on the C-ring allows for the rescue of antibacterial activity from certain resistance modifications that otherwise fully circumvent linezolid’s effect ⁹. Therefore, it is envisioned that novel oxazolidinones can be made with increased antibacterial activity, allowing for lower doses and reducing serious dose-dependent side effects ^2,10.

Though lacking the notoriety of their taxonomic relative Mycobacterium tuberculosis, non-tuberculous mycobacteria (NTM) such as M. abscessus represent a significant emerging disease threat ^18,19. The incidence of human infection with NTM, and especially M. abscessus, is rapidly rising, having overtaken tuberculosis in prevalence in most industrialized nations, including the United States ^20,21. A distinguishing feature of M. abscessus is the unique lipid-rich, mycolic acid containing cell wall, whose low permeability bestows intrinsic resistance to many antibiotics in clinical use ^20,22,23. Additionally, M. abscessus frequently displays increased expression of drug efflux pumps, drug neutralizing enzymes, and structural mutations to antibiotic targets that contribute to resistance ^20,24. Many drugs that are effective against M. tuberculosis and other mycobacteria such as rifampicin, isoniazid, pyrazinamide, and ethambutol are ineffective against M. abscessus due to their very high minimum inhibitory concentrations against this species that are intolerable in humans, and acquired and adaptive antibiotic resistance mechanisms encoded by its genome ^22,23. Given the lack of FDA-approved therapeutics to treat M. abscessus disease, current treatment regimens are often reliant on repurposing antibiotics approved for other indications ^20,22,23. These factors have resulted in a clinical treatment success rate of around 45%, contributing to M. abscessus’s reputation as a ‘clinical nightmare’ ²⁵. Despite the threat posed by increasing incidence of increasingly resistant M. abscessus disease, insufficient progress has been made to address the urgent need for new and effective pharmaceuticals to bridge this critical therapeutic gap²⁶.

To address this need, we have engaged in the development of novel oxazolidinone-containing antibiotics, derived from the benzodioxin moiety we previously reported in oxazolidinone T145 (3) (Figure 1) ^27–30. In the design of T145 we had three primary objectives. First, we sought to mimic the electronic nature of the phenyl ring in linezolid by changing the morpholino and fluorine substituents present in linezolid to a fused benzodioxin ring (Figure 1)^11,27. Secondly, we envisioned the conjugated 3-ring system present in 3 as conformationally constraining the molecule in comparison to linezolid, reducing the amount of free-rotating bonds, potentially contributing to an increased binding affinity ^2,15,27,28. Finally, we understood the benzodioxin moiety to be easily functionalized, enabling rationally informed structural extension capable of forming additional interactions ^2,11,27,28. Compound 3 has been demonstrated to possess superior activity to linezolid in inhibiting growth of the gram positive pathogenic bacteria Enterococcus faecalis, Staphylococcus aureus and M. tuberculosis ²⁸. Interestingly, in that report, we found that 3 demonstrated no activity less than 64 μg/mL MIC₉₀ against the NTM M. abscessus or Mycobacterium avium ²⁸.

Nonetheless, there is strong precedence for the use of oxazolidinones, including linezolid, for both in vitro inhibition and successful clinical treatment of M. abscessus^{16,22,23,31,32}. In an effort to enhance the effectiveness of our original oxazolidinone T145 (3) against M. abscessus, we have functionalized the benzodioxin aromatic ring system with a number of nitrogen containing functional groups. We found that regioselective introduction of an aldehyde group (Scheme 1) offered a reactive handle to introduce other functionality. We report the introduction of five functional groups in that position, the oxime-substituted T530 (12), the oxime ester-substituted T534 (13), the nitrile-substituted T538 (14), the amide substituted T537 (15), and the amidoxime-substituted T540 (16). These functional groups were chosen because they are not only hydrogen-bond acceptors, but also in the case of the amide and amidoxime present in 15 and 16, are also hydrogen-bond donors. We hypothesized that the characteristics of these groups would thus lead to the formation of additional interactions within the oxazolidinone binding site, leading to increased binding affinity, and improved antibacterial activity. Herein, we describe the synthesis and evaluate the antibacterial activity of these derivatives against M. abscessus.

Scheme 1.

Compound 11 was synthesized in 6 steps, using a modified and updated synthesis from our previous reports ^27,28. The double nucleophilic aromatic substitution of compounds 4 and 5 to produce our benzodioxin moiety is regioselective, yielding the psuedo-para isomer (6), likely owing to the resonance stabilizing effect of the hydroxyl group para to the aldehyde, making the meta hydroxyl more nucleophilic. The aldehyde (6) was protected by conversion to an acetal (7) in the second synthetic step, before being deprotected in acidic conditions in the final step of the synthesis of 11 ³³. The mechanism for formation of the oxazolidinone ring was altered from our previously syntheses to utilize the single step reaction from N-[(2S)-2-(Acetyloxy)-3-chloropropyl]acetamide as first reported by Perrault and colleagues ³⁴. Compound 11 was used as the precursor to make the other novel oxazolidinone derivatives discussed in this work.

The synthetic routes to make the five novel oxazolidinone compounds using compound 11 as starting material, are shown in Scheme 2. The aldehyde group of 11 was converted into five different functional groups: oxime 12, oxime ester 13, nitrile 14, amide 15, and amidoxime 16. All transformations proceeded in good yield, and low solubility of the products in most organic solvents allowed for chromatography-free purification.

Scheme 2.

The five novel compounds were tested for antibacterial activity against M. abscessus in accordance with the Clinical & Laboratory Standards Institute (CLSI) guidelines as described in the Supplemental Information ³⁵. The minimum inhibitory concentration (MIC) for linezolid against M. abscessus ATCC 19977 in Middlebrook 7H9 broth is 64 μg/mL ³⁵. The MIC of T538 (14) was >256 μg/mL, indicating antibacterial inactivity in this assay. T530 (12) and T534 (13), the oxime and derivatives respectively, was 128 μg/mL (Table 1). The amide-substituted T537 (15) showed improved activity (MIC = 64–128 μg/mL), but the most active was T540 (16) with a MIC of 16 μg/mL (Table 1); per the CLSI guidelines this MIC would be considered susceptible. As the MIC of 16 vs. M. abscessus ATCC 19977 is lower than that of linezolid (16 μg/mL as compared to 64 μg/mL), we have demonstrated that functionalization of the benzodioxin-substituted oxazolidinone scaffold can allow for the development of molecules that exhibit improved activity against M. abscessus.

Table 1.

Novel Derivative Structures and Antimicrobial Efficacy

Structure	Compound Name	MIC (μg/mL)
	T530	128
	T534	128
	T538	>256
	T537	64–128
	T540	16
	Linezolid	64

We theorize that the aldoxime functional group present in T540 (16), which is capable of forming hydrogen bond interactions as both a donor and acceptor, forms additional interactions with rRNA located within the binding site that do not occur with T145 (3). This is supported by the multiple published structures of linezolid (1) and other oxazolidinones bound to the bacterial ribosome, which demonstrate the presence of proximally located nucleotides with which linezolid forms no or very weak interactions, especially in the area of linezolid’s morpholine moiety ^2,12. Novel or stronger hydrogen bonding interactions with these nucleotides would contribute to an increased binding affinity to the 50S ribosomal subunit, leading to increased antibacterial activity.

Due to the potent activity of the amidoxime derivative 16 against M. abscessus, we sought to develop a shorter synthetic route. The original synthetic route to 16 was nine steps. We then devised a novel synthesis for compound 16 that takes only seven steps (Scheme 3). Notably, the reduction of the nitro group in the presence of a nitrile was performed chemoselectively using hypodiboric acid and a catalytic amount of 4,4’-bypyridine in DMF ³⁶. This route allowed for the more efficient synthesis of 16 with an overall yield of 22.0%.

Scheme 3.

Condensed Synthesis for Novel Nitrile Derivative

Additionally, in an attempt to form compound 18 in one step rather than three, the synthesis was performed starting with 3,4-dihydroxybenzonitrile (21) instead of 3,4-dihydroxybenzaldehyde (4). However, a mixture of isomers 18 and 22 were formed, as shown in Scheme 4, that could not be separated by conventional flash chromatography. Therefore, the route shown in Scheme 3 represents the more effective synthetic route to the compound with the best antibacterial activity.

Scheme 4.

Alternative route to 18

In conclusion, we have described the synthesis of five novel oxazolidinones compounds. The compounds were tested against M. abscessus and the MIC of one derivative, amidoxime 16, is lower than that of linezolid. Since linezolid is used to treat M. abscessus infection, amidoxime 16 presents a new opportunity to replace it to improve the overall treatment outcome.

Highlights:

• Concise, regioselective, chromatography-free synthesis yields compounds with activity against Mycobacterium abscessus.

• Site for structural extension to increase efficacy of benzodioxin-containing antibiotics is revealed.

• An amidoxime-containing derivative (16) is identified as a potent antibacterial against Mycobacterium abscessus.