The compound at the center of the story is 3-azidothiophene-2-carboxylic acid, abbreviated ACA, a small molecule previously shown to kill M. tuberculosis and M. bovis by simultaneously targeting two cell wall core assembly proteins called CpsA1 and CpsA2. This dual-targeting design matters because the two genes have partially redundant functions: single mutants survive, but a double deletion is lethal. For the bacterium to fully escape inhibition, mutations would have to occur in both genes at once, an event whose probability is the product of two individually rare mutation rates. Even so, earlier work found that resistant mutants could emerge at very low frequencies, roughly one in a hundred million cells, with modest increases in the minimum inhibitory concentration. That raised an important question: are there other, more accessible genetic routes that blunt the drug’s lethal effect without touching its direct protein targets?

To find out, the team constructed a transcription factor overexpression library in Mycobacterium bovis BCG, a close laboratory relative of the tuberculosis bacillus, and plated roughly ten billion cells on agar containing four and eight times the minimum inhibitory concentration of ACA. After a month of incubation, a single resistant colony appeared. The researchers named it BCG-R01 and measured its resistance: it tolerated three times the drug concentration that halted the wild-type strain, with a minimum inhibitory concentration of 9 micrograms per milliliter compared with 3 for the parent. The initial suspect was an overexpressed transcriptional regulator called CarD, but when the plasmid carrying it was extracted and reintroduced into fresh wild-type bacteria, no resistance followed. The real culprit had to be written into the genome itself.

Whole-genome sequencing of BCG-R01, validated by Sanger sequencing, revealed a single cytosine insertion at position 37 of the gene BCG_0932, known in M. tuberculosis as Rv0880. The insertion wrecked the reading frame, producing a truncated protein that stopped at position 46. The team christened the gene acasR, for ACA-sensitivity repressor. It encodes a member of the MarR family of transcriptional regulators, proteins that in bacteria commonly control stress responses, virulence, fatty acid metabolism and antibiotic resistance. Two other mutations turned up in the resistant clone, a frameshift in the PE_PGRS28 gene and a missense mutation in glycerol kinase, but overexpression experiments showed that neither of those genes influenced ACA sensitivity. The evidence pointed squarely at AcasR.

Complementary experiments sealed the case. Deleting acasR raised the minimum inhibitory concentration of ACA to 6 micrograms per milliliter, a two-fold increase, and the knockout strain grew significantly better than wild type in drug-containing medium. Restoring a functional copy of the gene reversed the phenotype, bringing the concentration back down to 2.5 micrograms per milliliter. Strikingly, the opposite manipulation had the opposite effect: strains engineered to overproduce AcasR became markedly more vulnerable to the drug. In other words, the more functional AcasR a mycobacterium carries, the more easily ACA kills it, and losing the regulator confers a measurable degree of protection.

With the genetic link established, the next challenge was to find what AcasR actually controls. Database predictions suggested two candidate targets, Rv1219c and Rv2590, the latter encoding a fatty acid CoA ligase called FadD9. Using electrophoretic mobility shift assays, in which purified AcasR protein is mixed with fluorescent or stained DNA fragments and run through a gel, the researchers observed clear shifted bands only with the promoter of fadD9, not with Rv1219c or with acasR’s own promoter. A specific 15-base-pair binding motif, TAATATATTTACTTT, sits between positions 90 and 76 upstream of the fadD9 start site, and replacing that motif with random sequence abolished AcasR binding entirely. Chromatin immunoprecipitation inside living bacteria confirmed the interaction in vivo, pulling down the fadD9 promoter with an AcasR-specific antibody from overexpressing strains but not from knockouts. Surprisingly, none of five additional genes predicted from the motif search bound AcasR, making fadD9 the sole functionally validated target under the tested conditions.

Reporter assays then established the direction of regulation. When the fadD9 promoter drove a lacZ reporter gene, expression rose significantly in an acasR-knockout background compared with wild type, demonstrating that AcasR is a repressor that clamps down on fadD9 transcription. The functional consequences were dramatic. Overexpressing fadD9 made bacteria more resistant to ACA, phenocopying the acasR mutation, while deleting fadD9 made them roughly twice as sensitive, with a minimum inhibitory concentration of just 1.5 micrograms per milliliter. Complementation rescued both phenotypes. Together these results sketch a clean regulatory axis: AcasR represses fadD9, and FadD9 levels determine how susceptible the cell is to ACA.

The connection to iron emerged from the team’s multi-omics analysis. Transcriptome sequencing of ACA-treated bacteria showed that the most strongly upregulated genes belonged to the mycobactin biosynthesis cluster, the Esx-3 siderophore secretion machinery, the mmpL4 and mmpS4 siderophore exporters, and the irtA and irtB iron importers, while iron storage genes bfrA and bfrB and numerous energy metabolism genes were switched off. This expression pattern closely mirrors what mycobacteria do when starved of iron. Lipidomics by liquid chromatography mass spectrometry confirmed the picture biochemically: iron-loaded mycobactin species surged as much as 30-fold under ACA treatment, while the respiratory quinones MK-8 and MK-9 dropped several-fold. Direct iron measurements using a ferrozine-based colorimetric assay showed that ACA caused iron to pile up in the insoluble cell envelope fraction while soluble cytoplasmic free ferrous iron plummeted. The most economical interpretation is that ACA blocks the transmembrane transport of iron-loaded carboxymycobactin, leaving iron stranded on the membrane surface while the cell interior starves.

The iron story also explains why fadD9 deletion sensitizes bacteria to the drug. Transcriptomics of the fadD9 knockout revealed downregulation of mycobactin synthesis genes including mbtE, mbtD and mbtC, and lipidomics showed a significant drop in iron-loaded mycobactin. When ACA was added to the fadD9-deficient strain, both membrane-associated and soluble iron fell even further than in wild type. Because fadD9 mutants already make less siderophore, the drug’s disruption of siderophore-mediated iron import strikes an already weakened supply chain, plunging the cell deeper into starvation. Interestingly, an engineered H37Ra strain lacking mbtB, an essential mycobactin biosynthesis enzyme, became resistant to ACA when the medium was supplemented with heme, demonstrating that mycobacteria can reroute their iron acquisition through the heme utilization pathway and thereby sidestep the drug’s effect on the mycobactin route.

The study’s implications reach beyond one experimental compound. Because fadD9 overexpression confers resistance, inhibitors of FadD9 could theoretically potentiate ACA’s bactericidal activity, and chemical families such as alkanoyl-AMS analogs, which target multiple FadD enzymes, offer a starting point for medicinal chemistry. The work also connects to broader themes in tuberculosis pharmacology: MarR family regulators such as Rv0678 are already notorious for causing bedaquiline and clofazimine resistance through efflux pump derepression, and Rv0880 itself was previously implicated in intrinsic bedaquiline tolerance. That a single regulatory axis ties iron homeostasis, fatty acid metabolism and drug susceptibility together underscores how deeply antibiotic action is entangled with basic bacterial physiology. The authors note open questions, including the precise mechanism by which FadD9 influences mycobactin gene expression, whether through altered lipid pools or a more direct metabolic coupling, and the need to test how ACA-induced iron starvation interacts with other anti-tuberculosis drugs under host-like conditions. For now, the finding provides a striking example of how a transcriptional regulator, a fatty acid ligase and a siderophore supply chain combine to decide whether a tuberculosis drug lives up to its promise.

Subject of Research: Transcriptional regulation of iron homeostasis and drug susceptibility in mycobacteria

Article Title: The AcasR-FadD9 regulatory axis modulates iron homeostasis and susceptibility to the anti-mycobacterial compound ACA

Article References: Huang, F., Wang, Z., Jiang, D., Li, S., He, Z.-G., & Zhang, H. (2026). The AcasR-FadD9 regulatory axis modulates iron homeostasis and susceptibility to the anti-mycobacterial compound ACA. Frontiers in Tuberculosis, 4, Article 1830528. https://doi.org/10.3389/ftubr.2026.1830528

Image Credits: AI Generated

DOI: 10.3389/ftubr.2026.1830528

Keywords: tuberculosis, ACA, AcasR, FadD9, iron homeostasis, siderophore, mycobactin, drug resistance, MarR regulator, Mycobacterium tuberculosis, CpsA1, CpsA2