Design and synthesis of water-soluble prodrugs of rifabutin for intraveneous administration

https://doi.org/10.1016/j.ejmech.2022.114515Get rights and content

Highlights

  • Design and synthesis of new water soluble C21-fonctionnalized rifabutin prodrugs.

  • Assessment of aqueous solubility, instability in murine and human plasma and activity on A. baumannii.

  • Proof of concept of bioequivalence to rifabutin in a mouse model.

Abstract

Acinetobacter baumannii is a gram-negative bacterium causing severe hospital-acquired infections such as bloodstream infections or pneumonia. Moreover, multidrug resistant A. baumannii becomes prevalent in many hospitals. Consequently, the World Health Organization made this bacterium a critical priority for the research and development of new antibiotics. Rifabutin, a semisynthetic product from the rifamycin class, was recently found to be very active in nutrient-limited eukaryotic cell culture medium against various A. baumannii strains, including extremely drug-resistant strains, with minimal inhibitory concentrations as low as 0.008 μg/mL. Moreover, this in vitro potency translates into in vivo efficacy. Thus, rifabutin appears to be an attractive novel antibiotic against A. baumannii. In this work, our objective was to design and synthetize rifabutin prodrugs with increased aqueous solubility to allow intraveneous use. Synthetic methodologies were developed to selectively functionalize the hydroxyl group in position 21 and to afford 17 prodrugs. We measured the water solubility of the prodrugs, the stability in human and mouse plasma and their antimicrobial activity against A. baumannii after incubation in human serum. Finally, a pharmacokinetic release study of rifabutin was performed in CD1 mice with three selected prodrugs as a proof of concept.

Introduction

Rifabutin (also known as LM 427 and Mycobutin®) is a spiro-piperidyl-rifamycin derived from rifamycin-S, belonging to the class of ansamycins. The antimicrobial activity of the rifamycins is based on their ability to penetrate the bacterial cell wall and to inhibit DNA-dependent RNA polymerase with a subsequent inhibition of transcription and protein synthesis. Rifabutin has a broad spectrum of antimicrobial activity. This includes activity against mycobacteria, a variety of Gram-positive and Gram-negative bacteria, Chlamydia trachomatis, and Toxoplasma gondii [1]. Escherichia coli and other Gram-negative enteric bacteria and nonfermenting Gram-negative bacilli are resistant to rifabutin at concentrations that can be readily achieved in blood [2]. It has been established that the drug's penetration through the bacteria cell wall explains the higher potency against Gram-positive compared to the Gram-negative despite the quite similar RNA polymerase inhibitory activities [3].

Acinetobacter baumannii is a Gram-negative pathogen causing severe hospital-acquired infections such as pneumonia and sepsis, with a fatality rate between 50 and 60% [4]. The World Health Organization categorized carbapenem resistant A. baumannii as a critical priority for the research and development of new antibiotics due to current limited effective therapeutic options [5]. In this context, the ReFRAME library, a collection of 12,000 compounds that have been marketed or have reached clinical development [6], was screened for in vitro activity against A. baumannii in nutrient-limited eukariotic cell culture medium to mimic the host environment. Rifabutin was identified as the most potent hit [7] and displayed minimal inhibitory concentrations (MICs) against A. baumannii strains as low as 0.008 μg/mL. Interestingly, MICs of rifampicin, similarly to all other rifamycin antibiotic tested, remained invariably high independently of the culture medium used. In studies of the mechanism of action, rifabutin was shown to be actively transported in A. baumannii through the TonB-dependent siderophore transporter FhuE [7]. More importantly, in vitro activity in iron-limited condition was shown to translate well into in vivo efficacy [7] and PK/PD studies revealed that whereas AUC/MIC is the primary driver for efficacy, Cmax/MIC is an important secondary driver and may be especially important not only for efficacy but to reduce chances of resistance development [8]. Thus, to optimize rifabutin exposures (AUC) and Cmax and to mininimize the intrinsic interpatient variability inherent to oral dosing [9], the intrinsic low acqueous solubility of rifabutin was overcome with a new intravenous (IV) formulation (BV100) [US20210077470A1]. BV100 is currently evaluated in three clinical phase I trial studies (NCT04636983, NCT05087069, NCT05086107) [[10], [11], [12]].

Alternatively, we aimed at synthesizing prodrugs capable of rapidly regenerating rifabutin in human plasma while demonstrating improved aqueous solubility. We report here the design, the synthesis and the physicochemical and pharmacokinetic characterization of rifabutin prodrugs for intravenous use. To date, only bisphosphonate rifabutin prodrugs have been studied to target bone infections caused by Staphyloccocus aureus and release the parent drug at the site of infections [13,14].

Two points of attachment were considered to introduce the promoiety on the rifabutin scaffold with three different chemical bonds: the imidazoline ring through regioselective acylation using either acyl chlorides or chloroformates to yield amides and carbamates, respectively [[15], [16], [17], [18]] or the hydroxyl group at the C-21 position through regioselective acylation using anhydrides to form esters [13,14,19].

As described in Scheme 1, rifabutin prodrugs amide 1, carbamate 2 and ester 3 were synthesized and their stability in human plasma and their MIC on A. baumannii were evaluated in vitro (Table 1).

Amide 1 had quantitatively released rifabutin within 6 h (Table 1). Unfortunately, this compound as well as other amide derivatives were found to be unstable under purification conditions or in the presence of water. Due to the overall poor chemical stability of this series, this functionalization was discarded. Carbamate 2 was more stable in aqueous solution but also in human plasma with a remaining percentage at 6 h of 86%. Nevertheless, a potent bacterial growth inhibition was observed, with a MIC of 0.06 μg/mL, which is in accordance with published N-substituted derivatives that maintained potency against S. aureus and Mycobacterium tuberculosis [14,15]. Therefore, carbamate 2 could not be considered as a true prodrug. Finally, ester 3 was also stable in plasma (99% remaining at 6 h) and did not show any A. baumannii growth inhibition at the tested concentrations (MIC >32 μg/mL). Indeed, the hydroxyl group at the C-21 position is mandatory for the interaction with bacterial RNA-polymerase and therefore for the growth inhibition activity [20]. As enzymatic conversion of prodrugs is dependent on accessibility of the scissile bond to the enzyme, we decided to pursue modifications of the C-21 position with spacers between the scissile bond and the rifabutin core. To improve the hydrolysis of these derivatives in plasma, we employed intramolecular spontaneous cyclization strategies and self-immolative linkers (Fig. 1). The hydroxyl group at C-21 was connected through an ester bound to a linker, which carries a nucleophilic atom (nitrogen or oxygen). Nucleophilicity of the latter is masked by an amide, an ester or a phosphate bound linked to the solubilizing group. Upon enzymatic deacylation of the nitrogen, spontaneous intramolecular cyclisation could result in rifabutin release (Type A rifabutin prodrugs, Fig. 1). We also considered the use of self-immolative linkers such as p-hydroxybenzylalcohol (PHBA) or “trimethyl lock” (TML) systems (Type B rifabutin prodrugs, Fig. 1) [[21], [22], [23]].

Section snippets

Synthesis of type A prodrugs

Type A rifabutin prodrugs were synthesized by coupling moieties carrying the solubilizing group through an amide bound to Gly-rifabutin (intermediate 6, Fig. 2) or Sar-rifabutin (intermediate 7, Fig. 2). The first strategy to increase water-solubility was to use l-amino acids as prodrug moieties. The second strategy was to incorporate a phopshate group via a glycolyl linker, and the third strategy was to introduce a dimethylglycine via the same linker. Three different series of type A prodrugs,

Human plasma stability

To investigate the stability of the prodrugs and their ability to release rifabutin, compounds were first incubated for 6 h, at 37 °C, in human plasma. The release of rifabutin was monitored by LC-MS. Prodrugs that showed a significant release of rifabutin were then tested, in duplicate, in a kinetic study, where the quantitative dosing of rifabutin was monitored by LC-MS/MS together with the remaining percentage of prodrug and key intermediates.

Conclusion

Rifabutin has recently been shown to have potent in vitro and in vivo activity against A. baumannii [7,43,44]. Given its limited oral bioavailability and low solubility, the development of water-soluble prodrugs of rifabutin would allow its intravenous administration at higher doses, which could result in increased AUC compared to oral administration at the same dose. This increased exposure would positively influence the AUC/MIC-dependent killing power of rifabutin, leading to increased

Materials and methods

Experiments with replicates were not independent except the procedure for MIC determination.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Birgit Schellhorn, Vincent Trebosc, Marc Gitzinger, Glenn E. Dale, Sergio Lociuro are Bioversys AG employees. Marc Gitzinger is a shareholder of Bioversys AG. Marilyne Bourotte and Olivier Defert are BioVersys SAS employees. Nicolas Willand is consultant for Bioversys AG The other authors declare that they have no competing interests.

Acknowledgment

This research was financially co-funded by European Union under the European Regional Development Fund (ERDF) and by the Hauts De France Regional Council (Contract n°NP0020070).

References (45)

  • M. Vaara

    Comparative activity of rifabutin and rifampicin against gram-negative bacteria that have damaged or defective outer membranes

    J. Antimicrob. Chemother.

    (1993)
  • M.-F. Lin et al.

    Antimicrobial resistance in acinetobacter baumannii: from bench to bedside

    World J. Clin. Cases WJCC

    (2014)
  • World Health Organization

    WHO publishes list of bacteria for which new antibiotics are urgently needed

  • J. Janes et al.

    The ReFRAME library as a comprehensive drug repurposing library and its application to the treatment of cryptosporidiosis

    Proc. Natl. Acad. Sci. U. S. A.

    (2018)
  • B. Luna et al.

    A nutrient-limited screen unmasks rifabutin hyperactivity for extensively drug-resistant acinetobacter baumannii

    Nat. Microbiol.

    (2020)
  • A. Muller et al.

    Pharmacokinetics and pharmacodynamics of BV100 in neutropenic mouse lung infection models

    ECCMID 2021 congress

    (2021)
  • M.H. Skinner et al.

    Pharmacokinetics of rifabutin

    Antimicrob. Agents Chemother.

    (1989)
  • Clinical trial to investigate the safety, tolerability and pharmacokinetics of BV100 in male subjects

  • Clinical Trial to Investigate the safety, tolerability and pharmacokinetics of BV100 in male subjects

  • Pharmacokinetics and safety of BV100 administered as single intravenous infusion to subjects with renal impairment

  • E. Dietrich et al.

    Phosphonated Rifamycins and Uses Thereof for the Prevention and Treatment of Bone and Joint Infections

    (2010)
  • E. Dietrich et al.

    Phosphonated Rifamycins and Uses Thereof for the Prevention and Treatment of Bone and Joint Infections

    (2011)
  • Cited by (5)

    View full text