“One-Two Punch”: Synergistic ß-Lactam Combinations for Mycobacterium abscessus and Target Redundancy in the Inhibition of Peptidoglycan Synthesis Enzymes

Abstract

Mycobacterium abscessus subsp. abscessus is one of the most difficult pathogens to treat and its incidence in disease is increasing. Dual β-lactam combinations act synergistically in vitro but are not widely employed in practice. A recent study shows that a combination of imipenem and ceftaroline significantly lowers the minimum inhibitory concentration of clinical isolates, despite both drugs targeting the same peptidoglycan synthesis enzymes. The underlying mechanism of this effect provides a basis for further investigations of dual β-lactam combinations in the treatment of M. abscessus subsp. abscessus, eventually leading to a clinical trial. Furthermore, dual β-lactam strategies may be explored for other difficult mycobacterial infections.

Interest in dual β-lactam treatment for Mycobacterium abscessus subsp. abscessus is growing. We summarize recent work that suggests drug target redundancy underlies in vitro efficacy and outline the implications this has for future studies for this difficult infection.

A recent study advances the notion of target redundancy underlying the synergism of a dual β-lactam combination against Mycobacterium abscessus subsp. abscessus, a rapidly growing nontuberculous mycobacterium (NTM) [1]. Mycobacterium abscessus consists of 3 subspecies: M. abscessus subsp. abscessus, M. abscessus subsp. bolletii, and M. abscessus subsp. massiliense [2]. The incidence of M. abscessus in disease is increasing and M. abscessus subsp. abscessus is among the most difficult pathogens to treat [3, 4]. Clinical disease commonly manifests as pulmonary infections or skin and soft tissue infections but can also include central nervous system, ocular, and disseminated infections [5]. Patients susceptible to M. abscessus pulmonary infection include those with underlying structural lung disease such as bronchiectasis, with or without cystic fibrosis (CF) and chronic obstructive pulmonary disease [4, 5]. Among patients with CF, M. abscessus pulmonary infection is associated with a decline in pulmonary function [6]. Additionally, hosts with compromised cell-mediated immunity are more likely to have disseminated M. abscessus infection [7].

Further complicating the tendency to infect medically complex hosts, M. abscessus possesses intrinsic and acquired mechanisms of antibiotic resistance [8]. Macrolide resistance is associated with higher rates of treatment failure and can be conferred by mutations in the 23S rRNA (rrl) gene or induction of the erm(41) gene [9–11]. Mycobacterium abscessus subsp. massiliense has a nonfunctional erm(41) gene and has a higher cure rate compared with the other subspecies [4]. Conversely, M. abscessus subsp. abscessus and M. abscessus subsp. bolletii isolates frequently harbor erm(41) and treatment for M. abscessus subsp. abscessus has a failure rate that is worse than those of multidrug-resistant tuberculosis (MDR-TB) and pulmonary infection with Mycobacterium avium complex (MAC) [4]. Resistance to multiple drug classes precludes a standardized therapy for M. abscessus and guidelines recommend designing a multidrug regimen based on in vitro susceptibility testing [9].

Although several β-lactam antibiotics were studied against isolates, only cefoxitin and imipenem are included in guideline recommendations for the treatment of M. abscessus [9, 12]. Story-Roller and colleagues [12] reviewed 35 studies and reported a wide range of β-lactam minimum inhibitory concentrations (MICs) against M. abscessus clinical isolates and nearly all studies included strains with cefoxitin or imipenem resistance. One major factor of β-lactam resistance in all mycobacteria is the presence of β-lactamases. The chromosomally encoded Bla_Mab is a serine class A β-lactamase with broad activity and relatively resistant to clavulanate; cefoxitin and imipenem are both slowly hydrolyzed by Bla_Mab (low k_cat/K_m), which contributes to their efficacy [13]. Possibly due to the insensitivity of Bla_Mab to clavulanate, β-lactamase inhibitors have not been widely employed in the treatment of M. abscessus. However, diazabicyclooctane (DBO) inhibitors (avibactam, relebactam, nacubactam, and zidebactam) demonstrate a restoration of β-lactam susceptibility and inhibit growth of M. abscessus [14–19].

Circumvention of Bla_Mab is only 1 challenge that must be addressed in using β-lactam antibiotics against M. abscessus. Like many mycobacteria, M. abscessus makes significant use of L,D-transpeptidases (Ldts) in peptidoglycan synthesis [20]. These enzymes catalyze a 3,3-crosslink between the peptidoglycan oligopeptides instead of the 4,3-crosslinks of the more widely recognized D,D-transpeptidases (also known as penicillin-binding proteins [PBPs]) that are the first-described β-lactam targets. Ldts of M. abscessus are resistant to penicillins but inhibited by cephalosporins and carbapenems [21]. Furthermore, in vitro dual β-lactam combinations act synergistically against M. abscessus ATCC 19977 and clinical isolates [21–23]. Notably, Pandey and colleagues [23] were able to demonstrate that ceftazidime in combination with ceftaroline or imipenem was effective in the treatment of THP-1 human macrophages infected with clinical isolates of M. abscessus, thus addressing an often-raised concern about the ability of β-lactams to get into the intracellular space.

With the above in mind, an investigation was undertaken to see if the growth of M. abscessus could be inhibited in cell-based assays by a combination of ceftaroline and imipenem and if the DBOs avibactam and relebactam were able to further potentiate the action of these agents [1]. The authors reasoned that combinations that could inhibit Ldts, PBPs, and Bla_Mab would result in lowered MICs compared with any 1 agent alone. Combinations were tested against a collection of M. abscessus subsp. abscessus clinical isolates. A fixed concentration of ceftaroline to imipenem markedly lowered MICs to a degree not previously seen in vitro with other dual β-lactam combinations (Figure 1) [1, 21–23]. Interestingly, the addition of relebactam did not enhance this effect, suggesting that the dual β-lactam combination was the predominant driver of the in vitro impact. Avibactam and relebactam only modestly increased ceftaroline activity and did not significantly improve imipenem activity [1].

Figure 1.

MIC distributions against 55 clinical isolates of Mycobacterium abscessus subsp. abscessus [1] Reprinted with permission. Abbreviation: MIC, minimum inhibitory concentration.

To provide a biochemical rationale for these observations, enzyme kinetic assays, timed mass spectrometry, and molecular modeling and docking with Bla_Mab and antibiotics were performed. Both ceftaroline and imipenem were relatively poor substrates for Bla_Mab as they had relatively low k_cat/K_m, with that of ceftaroline being nearly 10-fold lower than that of imipenem. Furthermore, a complex of ceftaroline and Bla_Mab was not captured on mass spectrometry. Avibactam was a more potent and efficient inhibitor of Bla_Mab than relebactam with nearly 500-fold lower K_{i app} and 1000-fold higher acylation rate (k₂/K). These kinetic findings would support the observation that avibactam lowered MICs better than relebactam. Molecular docking of each DBO into Bla_Mab provides further insight into the structural mechanism underlying these findings. Perhaps the most notable observation in the molecular docking is the steric hindrance between F237 of Bla_Mab and the piperidine group of relebactam [1].

This analysis also explored the action of ceftaroline and imipenem on the Ldts of M. abscessus, which would further add to the explanation of the efficacy behind their combination. Both ceftaroline and imipenem individually bound to Ldt_Mab1, Ldt_Mab2, Ldt_Mab4, Ldt_Mab5, and D,D-carboxypeptidase (Figure 2). When each of these enzymes were incubated with ceftaroline and imipenem together, preferential binding for imipenem binding was observed [1].

Figure 2.

Redundancy and interactions between imipenem, ceftaroline, avibactam, and relebactam and Ldt_Mab1 to Ldt_Mab5, D,D-carboxypeptidase, and Bla_Mab [1] Reprinted with permission. Abbreviation: Ldt, L,D-transpeptidase.

The target redundancy between imipenem and ceftaroline was unexpected and, at first, it may seem counterintuitive that it explains the synergistic effect of their combination for M. abscessus. After all, this can be contrasted with the concept behind a well-known synergistic dual β-lactam combination: ampicillin and ceftriaxone for Enterococcus faecalis endocarditis. In this combination, ceftriaxone inactivates PBP2 and PBP3 to complement the inactivation of PBP4 and PBP5 by ampicillin [24, 25]. So how is a synergistic effect against M. abscessus explained when the 2 agents bind to the same enzyme targets?

Two hypotheses are proposed from the findings. Primarily, alteration of the peptidoglycan composition can be directly attributable to the number and sequence of cell wall synthesis enzymes inactivated. After imipenem has been bound and turned over by the Ldts or D,D-carboxypeptidase, ceftaroline may bind next, inactivate the enzyme, continue to disrupt peptidoglycan synthesis, and increase the permeability of the cell wall. Alternatively, ligand-induced conformational changes in the enzyme targets could permit secondary binding of another β-lactam. This latter hypothesis is supported by the authors’ in silico analyses showing that it is necessary for the active site cavity of Ldt_Mab2 to be flexible in order to accommodate ceftaroline; after initial preferential binding of imipenem and subsequent turnover, the cavity may be left in a conformation that is more favorable for ceftaroline binding [1].

Although relebactam did not significantly add to the combination of imipenem and ceftaroline for M. abscessus, this does not necessarily rule out a DBO role in therapy. Avibactam has been shown to have some inhibitory activity of PBPs and Ldts [26, 27]. As a result, we maintain that another DBO paired with an appropriate β-lactam partner may be effective through the similar mechanism of peptidoglycan synthesis enzyme target redundancy in conjunction with inhibition of Bla_Mab.

By demonstrating in vitro efficacy and providing a biochemical rationale, this analysis enhances the understanding of therapeutic combinations against M. abscessus through targeting peptidoglycan synthesis. In vivo studies will be needed to confirm efficacy of a dual β-lactam strategy. Using a recently developed mouse model of pulmonary M. abscessus infection, Story-Roller and colleagues demonstrated synergy of 4 dual β-lactam combinations delivered via aerosol route [28, 29]. Further animal model and clinical trials will be needed to define optimal combinations, dosing regimens, and time course, but this work points toward a promising direction to address this difficult infection in dire need of options. In the case of CF, effective therapies for M. abscessus could further translate into a chance to slow down an infection-related decline in pulmonary function. Indeed, clinicians may already seek to try a dual β-lactam strategy with their individual patients desperate for a cure for their M. abscessus infections; in these cases, we encourage clinicians to track and report outcomes as we await a more robust evidence base.

The results of this work also have significant implications for other difficult mycobacterial infections. The use of β-lactams and β-lactamase inhibitors for mycobacterial infections may be of particular interest in people living with human immunodeficiency virus (PLWH) where drug–drug interactions can be a concern. The NTM infections of importance to PWLH include MAC and Mycobacterium kansasii. However, Mycobacterium tuberculosis is the leading cause of infectious mortality in PLWH worldwide [30, 31]. The use of newer agents in regimens for MDR-TB can be limited by toxicity, drug–drug interactions, and cost [32]. Mycobacterium tuberculosis also makes use of PBPs and Ldts and has a class A β-lactamase, BlaC, that is inhibited by clavulanate [33, 34]. For M. tuberculosis, β-lactam antibiotics have played an even smaller treatment role than with M. abscessus: meropenem or imipenem combined with clavulanate (via co-formulation with amoxicillin) is a treatment option for MDR-TB when more preferred therapies are not feasible [35, 36]. The data to support this notion are largely observational, except for small, randomized trials with rifampin-susceptible M. tuberculosis demonstrating a decrease in sputum mycobacterial load similar to standard therapy [36–39]. Could the insights gained with M. abscessus also be applied to advance the role of β-lactams in M. tuberculosis treatment? In other words, might target redundancy between an optimized combination of agents also be the key to synergistic killing of M. tuberculosis via peptidoglycan synthesis inhibition? If such a combination can be designed and demonstrated to be effective, it would provide an opportunity to use this safe and comparatively well-tolerated group of drugs for pathogens that continue to plague humanity.

Notes

Acknowledgments. We acknowledge the contributions of Magdalena A. Taracila, Tracey Bonfield, Christopher R. Bethel, Melissa D. Barnes, Suresh Selvaraju, Ayman M. Abdelhamed, and Shannon H. Kasperbauer who were authors on the study in reference [1].

Disclaimer. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Department of Veterans Affairs.

Potential conflicts of interest. S. G. K. reports personal fees from Rockpointe CME (speaker fee for educational activity on NTM infections), outside the submitted work. C. L. D. reports grants and personal fees from Insmed, personal fees from Paratek, personal fees from Spero, personal fees from AN2, personal fees from Matinas, and grants from BugWorks, outside the submitted work. R. A. B. reports grants from the National Institutes of Health/National Institute of Allergy and Infectious Diseases, grants from the VA Merit Review Program, grants from VenatoRx, grants from Merck, grants from Entasis, grants from Wockhardt, and personal fees from Merck SAB, outside the submitted work; in addition, R. A. B. has 2 patents issued unrelated to this work and a provisional patent is planned for submission. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.