Toxin-Driven Antibiotic Resistance in Mycobacterium abscessus Circumvented by DNA Gyrase and RNA Polymerase Inhibitors

Mycobacterium abscessus is among the most stubborn bacterial pathogens in clinical medicine, resisting clearance even after months of aggressive combination therapy. Now researchers at Rutgers Health—Robert Wood Johnson Medical School have reported a strategy that could tip the balance back toward the patient. In a study published in Frontiers in Tuberculosis, Nikita Amin, Eduardo Troian and Nancy Woychik show that a toxin-antitoxin system called VapBC5, which clinical strains of M. abscessus can acquire and which substantially amplifies the bacterium’s already formidable drug resistance, leaves a characteristic weak point: antibiotics that attack DNA replication or transcription rather than the cell wall or the ribosome retain their killing power even when the toxin is fully active.

The finding rests on a detailed dissection of what the VapC5 toxin actually does inside the bacterium. VapBC5 belongs to the type II family of toxin-antitoxin systems, in which a labile antitoxin normally keeps its cognate toxin in check; under stress, the antitoxin is degraded and the toxin is freed to act on its cellular targets. When VapC5 is released, it specifically cleaves a single transfer RNA species, tRNASerGCA, depleting the cell’s capacity to translate serine codons. This imposes growth arrest and, crucially, reprograms protein synthesis: transcripts with low serine codon content, including many encoding ribosomal proteins, are preferentially made. The result is a bacterium that stops dividing and stockpiles ribosomes, a physiological state that blunts the lethality of antibiotics targeting cell-wall biosynthesis, such as cefoxitin, and those targeting protein translation, such as amikacin and tedizolid.

In the new work, the team set out to map the toxin’s influence more comprehensively than before. Their earlier study had examined only five well-characterized members of the WhiB7 regulon, the transcriptional program governed by a conserved mycobacterial transcription factor that underpins much of M. abscessus intrinsic resistance. Drawing on datasets from earlier studies that defined 127 WhiB7-responsive transcripts and, through chromatin immunoprecipitation followed by sequencing, 73 genes whose promoters are directly bound by WhiB7, the researchers asked how many of these VapC5 actually activates. The answer was selective rather than sweeping: of the 127 WhiB7-responsive transcripts, VapC5 increased the steady-state levels of 90, and of the 73 direct WhiB7 targets, nearly all were upregulated in wild-type cells. VapC5, in other words, does not switch on the entire regulon indiscriminately but induces a distinct, largely WhiB7-dependent subset of resistance genes.

To separate WhiB7-dependent effects from those the toxin exerts on its own, the researchers deleted the whiB7 gene using oligonucleotide-mediated recombineering followed by Bxb1 integrase targeting, then expressed VapC5 in the deletion background and profiled the global transcriptional response by RNA sequencing. Growth arrest persisted without WhiB7, confirming that the toxin’s primary insult, tRNA cleavage, is independent of the regulator. But the protective program collapsed. The growth-arrested ΔwhiB7 cells no longer survived exposure to cefoxitin or amikacin in persister assays, and the characteristic upregulation of ribosomal protein transcripts vanished. Functional enrichment analysis of the remaining differentially expressed genes pointed instead to two themes: DNA repair and lipid biosynthesis.

The most striking transcriptional signature in WhiB7-deficient cells came from an unexpected corner of the genome. Eight of the ten most strongly induced transcripts, ranging from 7.6 to 4.7 log2 fold change, encoded hypothetical proteins, and the majority clustered within a single genomic region spanning MAB_0222c through MAB_0233, a locus that also encodes a putative phage integrase and a phage tail tape-measure protein. Only one of these genes, MAB_0224, was significantly induced in wild-type cells, and far more weakly. The concomitant activation of SOS and DNA damage response pathways lent weight to the authors’ interpretation: DNA damage is a classic trigger of prophage induction, and M. abscessus strains commonly carry prophage-like sequences and mobile genetic elements in their accessory genomes. The researchers suggest this locus may reside within such an element and that its activation contributes to the maladaptive state that leaves ΔwhiB7 cells unable to survive cell-wall or translation-targeting antibiotics.

The stress-response regulator SigH added a further layer. SigH responds to heat, oxidative and nitric oxide stress, all common triggers of toxin-antitoxin activity, and had previously been shown to be activated by WhiB7 and to induce 33 genes outside the WhiB7 regulon. When VapC5 was expressed in the ΔwhiB7 background, the SigH response was significantly dampened, indicating that WhiB7 is required for full activation of this stress signature as well. Taken together, the deletion experiments recast VapC5 not as a standalone resistance factor but as a physiological modulator whose protective output depends heavily on the WhiB7 channel: without it, the same toxin pushes the cell toward a damaged, phage-tinged stress state rather than a drug-tolerant one.

With the mechanism in hand, the team turned to the therapeutic question. Because VapC5 protects M. abscessus specifically from drugs that target cell-wall biosynthesis or protein synthesis, they reasoned that antibiotics with entirely different mechanisms of action should retain activity. The candidate pool was constrained: most M. abscessus treatment options hit the ribosome or the cell wall, and several alternatives fell away for practical reasons. The DNA gyrase inhibitor evybactin is prohibitively expensive with long lead times; clofazimine showed variable minimum inhibitory concentrations depending on reading day and medium composition, as did the ATP synthase inhibitor bedaquiline. Two drugs met the criteria: ciprofloxacin, a fluoroquinolone that targets DNA gyrase, and rifabutin, a rifamycin that targets the beta subunit of RNA polymerase.

Both antibiotics killed VapC5-expressing cells as efficiently as controls. The researchers first determined the minimum inhibitory concentration of each drug as 2 micrograms per milliliter in the ATCC 19977 reference strain, then exposed cultures with or without toxin induction to ten times the MIC of ciprofloxacin or fifty times the MIC of rifabutin for up to 48 hours, using concentrations well above the MIC to distinguish true persistence from transient resistance. Time-kill curves for induced and uninduced cells essentially overlapped for both drugs, and viable cell recovery showed no increase in persisters. This stands in sharp contrast to the reduced killing and enhanced persister formation the same toxin produces against cefoxitin, amikacin and tedizolid. The authors note that rifamycins are generally weak against M. abscessus because they are vulnerable to inactivation by oxidation and by the ADP-ribosyltransferase Arr, but optimized rifabutin analogs that evade Arr have recently shown superior efficacy in a mouse model, keeping the class in play.

The rifabutin result carried an additional dividend for combination therapy. Clarithromycin is frequently chosen for initial treatment of M. abscessus infections, but its efficacy is often undermined by the unmutated, inducible erm41 ribosomal methylase gene carried by many strains, including the reference strain used in these experiments. Previous work from the Dick laboratory showed that rifabutin, by inhibiting RNA polymerase, suppresses transcription of both whiB7 and erm41, restoring clarithromycin sensitivity in otherwise resistant strains and motivating proposals to test rifabutin as a frontline partner in therapeutic trials. Because VapC5 expression itself drives whiB7 upregulation, the Rutgers team asked whether the toxin might interfere with this synergy. Quantitative reverse-transcription PCR showed no statistically significant difference in erm41 levels between VapC5-expressing and control cells treated with subinhibitory rifabutin, and whiB7 transcripts likewise remained low under treatment. The rifabutin-clarithromycin synergy, in other words, survives the toxin.

The authors propose that the findings could translate into a practical risk-stratification step. Screening clinical isolates for acquisition of the VapBC5 system, by PCR rather than the DNA sequencing needed to detect erm41 status or rrl mutations associated with clarithromycin resistance, could direct clinicians toward antibiotics whose targets lie outside the toxin’s protective reach. The evidence so far is in vitro, established in the ATCC 19977 reference strain under controlled conditions, and the researchers frame ciprofloxacin and optimized rifabutin analogs as a starting point for further study rather than an immediate change to treatment guidelines. Even so, the study sharpens the picture of how accessory genome elements shape drug tolerance in one of the hardest-to-treat bacterial pathogens, and it illustrates a broader principle: understanding the mechanism by which a resistance factor works can reveal which existing drugs it cannot touch.

Subject of Research: How the VapC5 toxin-antitoxin system reshapes antibiotic resistance in Mycobacterium abscessus and which drug classes retain efficacy

Article Title: Rationally tailored antibiotic selection circumvents toxin-mediated evasion of killing in Mycobacterium abscessus

Article References: Amin, N. H., Troian, E. A., & Woychik, N. A. (2026). Rationally tailored antibiotic selection circumvents toxin-mediated evasion of killing in Mycobacterium abscessus. Frontiers in Tuberculosis, 4, 1921029. https://doi.org/10.3389/ftubr.2026.1921029

Image Credits: AI Generated

DOI: 10.3389/ftubr.2026.1921029

Keywords: Mycobacterium abscessus, toxin-antitoxin system, VapC5, WhiB7, antibiotic resistance, persister cells, ciprofloxacin, rifabutin, DNA gyrase, RNA polymerase, tRNA cleavage, non-tuberculous mycobacteria