Emerging enterococcus pore-forming toxins with MHC/HLA-I as receptors.

Enterococci are a part of human microbiota and a leading cause of multidrug resistant infections. Here, we identify a family of Enterococcus pore-forming toxins (Epxs) in E. faecalis, E. faecium, and E. hirae strains isolated across the globe. Structural studies reveal that Epxs form a branch of ß-barrel pore-forming toxins with a ß-barrel protrusion (designated the top domain) sitting atop the cap domain. Through a genome-wide CRISPR-Cas9 screen, we identify human leukocyte antigen class I (HLA-I) complex as a receptor for two members (Epx2 and Epx3), which preferentially recognize human HLA-I and homologous MHC-I of equine, bovine, and porcine, but not murine, origin. Interferon exposure, which stimulates MHC-I expression, sensitizes human cells and intestinal organoids to Epx2 and Epx3 toxicity. Co-culture with Epx2-harboring E. faecium damages human peripheral blood mononuclear cells and intestinal organoids, and this toxicity is neutralized by an Epx2 antibody, demonstrating the toxin-mediated virulence of Epx-carrying Enterococcus.

Genome-wide gene expression tuning reveals diverse vulnerabilities of M. tuberculosis.

Antibacterial agents target the products of essential genes but rarely achieve complete target inhibition. Thus, the all-or-none definition of essentiality afforded by traditional genetic approaches fails to discern the most attractive bacterial targets: those whose incomplete inhibition results in major fitness costs. In contrast, gene "vulnerability" is a continuous, quantifiable trait that relates the magnitude of gene inhibition to the effect on bacterial fitness. We developed a CRISPR interference-based functional genomics method to systematically titrate gene expression in Mycobacterium tuberculosis (Mtb) and monitor fitness outcomes. We identified highly vulnerable genes in various processes, including novel targets unexplored for drug discovery. Equally important, we identified invulnerable essential genes, potentially explaining failed drug discovery efforts. Comparison of vulnerability between the reference and a hypervirulent Mtb isolate revealed incomplete conservation of vulnerability and that differential vulnerability can predict differential antibacterial susceptibility. Our results quantitatively redefine essential bacterial processes and identify high-value targets for drug development.

A selective antibiotic for Lyme disease.

Lyme disease is on the rise. Caused by a spirochete Borreliella burgdorferi, it affects an estimated 500,000 people in the United States alone. The antibiotics currently used to treat Lyme disease are broad spectrum, damage the microbiome, and select for resistance in non-target bacteria. We therefore sought to identify a compound acting selectively against B. burgdorferi. A screen of soil micro-organisms revealed a compound highly selective against spirochetes, including B. burgdorferi. Unexpectedly, this compound was determined to be hygromycin A, a known antimicrobial produced by Streptomyces hygroscopicus. Hygromycin A targets the ribosomes and is taken up by B. burgdorferi, explaining its selectivity. Hygromycin A cleared the B. burgdorferi infection in mice, including animals that ingested the compound in a bait, and was less disruptive to the fecal microbiome than clinically relevant antibiotics. This selective antibiotic holds the promise of providing a better therapeutic for Lyme disease and eradicating it in the environment.

A Dual-Mechanism Antibiotic Kills Gram-Negative Bacteria and Avoids Drug Resistance.

The rise of antibiotic resistance and declining discovery of new antibiotics has created a global health crisis. Of particular concern, no new antibiotic classes have been approved for treating Gram-negative pathogens in decades. Here, we characterize a compound, SCH-79797, that kills both Gram-negative and Gram-positive bacteria through a unique dual-targeting mechanism of action (MoA) with undetectably low resistance frequencies. To characterize its MoA, we combined quantitative imaging, proteomic, genetic, metabolomic, and cell-based assays. This pipeline demonstrates that SCH-79797 has two independent cellular targets, folate metabolism and bacterial membrane integrity, and outperforms combination treatments in killing methicillin-resistant Staphylococcus aureus (MRSA) persisters. Building on the molecular core of SCH-79797, we developed a derivative, Irresistin-16, with increased potency and showed its efficacy against Neisseria gonorrhoeae in a mouse vaginal infection model. This promising antibiotic lead suggests that combining multiple MoAs onto a single chemical scaffold may be an underappreciated approach to targeting challenging bacterial pathogens.

Crystal Structures of Membrane Transporter MmpL3, an Anti-TB Drug Target.

Despite intensive efforts to discover highly effective treatments to eradicate tuberculosis (TB), it remains as a major threat to global human health. For this reason, new TB drugs directed toward new targets are highly coveted. MmpLs (Mycobacterial membrane proteins Large), which play crucial roles in transporting lipids, polymers and immunomodulators and which also extrude therapeutic drugs, are among the most important therapeutic drug targets to emerge in recent times. Here, crystal structures of mycobacterial MmpL3 alone and in complex with four TB drug candidates, including SQ109 (in Phase 2b-3 clinical trials), are reported. MmpL3 consists of a periplasmic pore domain and a twelve-helix transmembrane domain. Two Asp-Tyr pairs centrally located in this domain appear to be key facilitators of proton-translocation. SQ109, AU1235, ICA38, and rimonabant bind inside the transmembrane region and disrupt these Asp-Tyr pairs. This structural data will greatly advance the development of MmpL3 inhibitors as new TB drugs.

Blood culture-free ultra-rapid antimicrobial susceptibility testing.

Treatment assessment and patient outcome for sepsis depend predominantly on the timely administration of appropriate antibiotics1-3. However, the clinical protocols used to stratify and select patient-specific optimal therapy are extremely slow4. In particular, the major hurdle in performing rapid antimicrobial susceptibility testing (AST) remains in the lengthy blood culture procedure, which has long been considered unavoidable due to the limited number of pathogens present in the patient's blood. Here we describe an ultra-rapid AST method that bypasses the need for traditional blood culture, thereby demonstrating potential to reduce the turnaround time of reporting drug susceptibility profiles by more than 40-60 h compared with hospital AST workflows. Introducing a synthetic beta-2-glycoprotein I peptide, a broad range of microbial pathogens are selectively recovered from whole blood, subjected to species identification or instantly proliferated and phenotypically evaluated for various drug conditions using a low-inoculum AST chip. The platform was clinically evaluated by the enrolment of 190 hospitalized patients suspected of having infection, achieving 100% match in species identification. Among the eight positive cases, six clinical isolates were retrospectively tested for AST showing an overall categorical agreement of 94.90% with an average theoretical turnaround time of 13 ± 2.53 h starting from initial blood processing.

A novel antibiotic class targeting the lipopolysaccharide transporter.

Carbapenem-resistant Acinetobacter baumannii (CRAB) has emerged as a major global pathogen with limited treatment options1. No new antibiotic chemical class with activity against A. baumannii has reached patients in over 50 years1. Here we report the identification and optimization of tethered macrocyclic peptide (MCP) antibiotics with potent antibacterial activity against CRAB. The mechanism of action of this molecule class involves blocking the transport of bacterial lipopolysaccharide from the inner membrane to its destination on the outer membrane, through inhibition of the LptB2FGC complex. A clinical candidate derived from the MCP class, zosurabalpin (RG6006), effectively treats highly drug-resistant contemporary isolates of CRAB both in vitro and in mouse models of infection, overcoming existing antibiotic resistance mechanisms. This chemical class represents a promising treatment paradigm for patients with invasive infections due to CRAB, for whom current treatment options are inadequate, and additionally identifies LptB2FGC as a tractable target for antimicrobial drug development.

Discovery of a structural class of antibiotics with explainable deep learning.

The discovery of novel structural classes of antibiotics is urgently needed to address the ongoing antibiotic resistance crisis1-9. Deep learning approaches have aided in exploring chemical spaces1,10-15; these typically use black box models and do not provide chemical insights. Here we reasoned that the chemical substructures associated with antibiotic activity learned by neural network models can be identified and used to predict structural classes of antibiotics. We tested this hypothesis by developing an explainable, substructure-based approach for the efficient, deep learning-guided exploration of chemical spaces. We determined the antibiotic activities and human cell cytotoxicity profiles of 39,312 compounds and applied ensembles of graph neural networks to predict antibiotic activity and cytotoxicity for 12,076,365 compounds. Using explainable graph algorithms, we identified substructure-based rationales for compounds with high predicted antibiotic activity and low predicted cytotoxicity. We empirically tested 283 compounds and found that compounds exhibiting antibiotic activity against Staphylococcus aureus were enriched in putative structural classes arising from rationales. Of these structural classes of compounds, one is selective against methicillin-resistant S. aureus (MRSA) and vancomycin-resistant enterococci, evades substantial resistance, and reduces bacterial titres in mouse models of MRSA skin and systemic thigh infection. Our approach enables the deep learning-guided discovery of structural classes of antibiotics and demonstrates that machine learning models in drug discovery can be explainable, providing insights into the chemical substructures that underlie selective antibiotic activity.

Streptomyces umbrella toxin particles block hyphal growth of competing species.

Streptomyces are a genus of ubiquitous soil bacteria from which the majority of clinically utilized antibiotics derive1. The production of these antibacterial molecules reflects the relentless competition Streptomyces engage in with other bacteria, including other Streptomyces species1,2. Here we show that in addition to small-molecule antibiotics, Streptomyces produce and secrete antibacterial protein complexes that feature a large, degenerate repeat-containing polymorphic toxin protein. A cryo-electron microscopy structure of these particles reveals an extended stalk topped by a ringed crown comprising the toxin repeats scaffolding five lectin-tipped spokes, which led us to name them umbrella particles. Streptomyces coelicolor encodes three umbrella particles with distinct toxin and lectin composition. Notably, supernatant containing these toxins specifically and potently inhibits the growth of select Streptomyces species from among a diverse collection of bacteria screened. For one target, Streptomyces griseus, inhibition relies on a single toxin and that intoxication manifests as rapid cessation of vegetative hyphal growth. Our data show that Streptomyces umbrella particles mediate competition among vegetative mycelia of related species, a function distinct from small-molecule antibiotics, which are produced at the onset of reproductive growth and act broadly3,4. Sequence analyses suggest that this role of umbrella particles extends beyond Streptomyces, as we identified umbrella loci in nearly 1,000 species across Actinobacteria.

Mycobacterium abscessus HelR interacts with RNA polymerase to confer intrinsic rifamycin resistance.

Rifampicin (RIF), the frontline drug against M. tuberculosis, is completely ineffective against M. abscessus, partially due to the presence of an ADP-ribosyltransferase (Arr) that inactivates RIF. Using RNA-seq, we show that exposure of M. abscessus to sublethal doses of RIF and Rifabutin (RBT), a close analog of RIF, results in an ∼25-fold upregulation of Mab_helR in laboratory and clinical isolates. An isogenic deletion in Mab_helR results in RIF/RBT hypersensitivity, and overexpression of Mab_helR confers RIF tolerance in M. tuberculosis. We demonstrate an increased HelR-RNAP association in RIF-exposed bacteria and a MabHelR-mediated dissociation of RNAP from stalled initiation complexes in vitro. Finally, we show that the tip of the PCh-loop of Mab_helR, present in proximity to RIF, is critical for conferring RIF resistance but dispensable for dissociation of stalled RNAP complexes, suggesting that HelR-mediated RIF resistance requires a step in addition to displacement of RIF-stalled RNAP.

Unraveling the biology of a fungal meningitis pathogen using chemical genetics.

The fungal meningitis pathogen Cryptococcus neoformans is a central driver of mortality in HIV/AIDS. We report a genome-scale chemical genetic data map for this pathogen that quantifies the impact of 439 small-molecule challenges on 1,448 gene knockouts. We identified chemical phenotypes for 83% of mutants screened and at least one genetic response for each compound. C. neoformans chemical-genetic responses are largely distinct from orthologous published profiles of Saccharomyces cerevisiae, demonstrating the importance of pathogen-centered studies. We used the chemical-genetic matrix to predict novel pathogenicity genes, infer compound mode of action, and to develop an algorithm, O2M, that predicts antifungal synergies. These predictions were experimentally validated, thereby identifying virulence genes, a molecule that triggers G2/M arrest and inhibits the Cdc25 phosphatase, and many compounds that synergize with the antifungal drug fluconazole. Our work establishes a chemical-genetic foundation for approaching an infection responsible for greater than one-third of AIDS-related deaths.

Porin-independent accumulation in Pseudomonas enables antibiotic discovery.

Gram-negative antibiotic development has been hindered by a poor understanding of the types of compounds that can accumulate within these bacteria1,2. The presence of efflux pumps and substrate-specific outer-membrane porins in Pseudomonas aeruginosa renders this pathogen particularly challenging3. As a result, there are few antibiotic options for P. aeruginosa infections4 and its many porins have made the prospect of discovering general accumulation guidelines seem unlikely5. Here we assess the whole-cell accumulation of 345 diverse compounds in P. aeruginosa and Escherichia coli. Although certain positively charged compounds permeate both bacterial species, P. aeruginosa is more restrictive compared to E. coli. Computational analysis identified distinct physicochemical properties of small molecules that specifically correlate with P. aeruginosa accumulation, such as formal charge, positive polar surface area and hydrogen bond donor surface area. Mode of uptake studies revealed that most small molecules permeate P. aeruginosa using a porin-independent pathway, thus enabling discovery of general P. aeruginosa accumulation trends with important implications for future antibiotic development. Retrospective antibiotic examples confirmed these trends and these discoveries were then applied to expand the spectrum of activity of a gram-positive-only antibiotic, fusidic acid, into a version that demonstrates a dramatic improvement in antibacterial activity against P. aeruginosa. We anticipate that these discoveries will facilitate the design and development of high-permeating antipseudomonals.

Tuning sterol extraction kinetics yields a renal-sparing polyene antifungal.

Decades of previous efforts to develop renal-sparing polyene antifungals were misguided by the classic membrane permeabilization model1. Recently, the clinically vital but also highly renal-toxic small-molecule natural product amphotericin B was instead found to kill fungi primarily by forming extramembraneous sponge-like aggregates that extract ergosterol from lipid bilayers2-6. Here we show that rapid and selective extraction of fungal ergosterol can yield potent and renal-sparing polyene antifungals. Cholesterol extraction was found to drive the toxicity of amphotericin B to human renal cells. Our examination of high-resolution structures of amphotericin B sponges in sterol-free and sterol-bound states guided us to a promising structural derivative that does not bind cholesterol and is thus renal sparing. This derivative was also less potent because it extracts ergosterol more slowly. Selective acceleration of ergosterol extraction with a second structural modification yielded a new polyene, AM-2-19, that is renal sparing in mice and primary human renal cells, potent against hundreds of pathogenic fungal strains, resistance evasive following serial passage in vitro and highly efficacious in animal models of invasive fungal infections. Thus, rational tuning of the dynamics of interactions between small molecules may lead to better treatments for fungal infections that still kill millions of people annually7,8 and potentially other resistance-evasive antimicrobials, including those that have recently been shown to operate through supramolecular structures that target specific lipids9.

Structural and mechanistic insights into fungal ß-1,3-glucan synthase FKS1.

The membrane-integrated synthase FKS is involved in the biosynthesis of ß-1,3-glucan, the core component of the fungal cell wall1,2. FKS is the target of widely prescribed antifungal drugs, including echinocandin and ibrexafungerp3,4. Unfortunately, the mechanism of action of FKS remains enigmatic and this has hampered development of more effective medicines targeting the enzyme. Here we present the cryo-electron microscopy structures of Saccharomyces cerevisiae FKS1 and the echinocandin-resistant mutant FKS1(S643P). These structures reveal the active site of the enzyme at the membrane-cytoplasm interface and a glucan translocation path spanning the membrane bilayer. Multiple bound lipids and notable membrane distortions are observed in the FKS1 structures, suggesting active FKS1-membrane interactions. Echinocandin-resistant mutations are clustered at a region near TM5-6 and TM8 of FKS1. The structure of FKS1(S643P) reveals altered lipid arrangements in this region, suggesting a drug-resistant mechanism of the mutant enzyme. The structures, the catalytic mechanism and the molecular insights into drug-resistant mutations of FKS1 revealed in this study advance the mechanistic understanding of fungal ß-1,3-glucan biosynthesis and establish a foundation for developing new antifungal drugs by targeting FKS.

Antibiotic combinations reduce Staphylococcus aureus clearance.

The spread of antibiotic resistance is attracting increased attention to combination-based treatments. Although drug combinations have been studied extensively for their effects on bacterial growth1-11, much less is known about their effects on bacterial long-term clearance, especially at cidal, clinically relevant concentrations12-14. Here, using en masse microplating and automated image analysis, we systematically quantify Staphylococcus aureus survival during prolonged exposure to pairwise and higher-order cidal drug combinations. By quantifying growth inhibition, early killing and longer-term population clearance by all pairs of 14 antibiotics, we find that clearance interactions are qualitatively different, often showing reciprocal suppression whereby the efficacy of the drug mixture is weaker than any of the individual drugs alone. Furthermore, in contrast to growth inhibition6-10 and early killing, clearance efficacy decreases rather than increases as more drugs are added. However, specific drugs targeting non-growing persisters15-17 circumvent these suppressive effects. Competition experiments show that reciprocal suppressive drug combinations select against resistance to any of the individual drugs, even counteracting methicillin-resistant Staphylococcus aureus both in vitro and in a Galleria mellonella larva model. As a consequence, adding a ß-lactamase inhibitor that is commonly used to potentiate treatment against ß-lactam-resistant strains can reduce rather than increase treatment efficacy. Together, these results underscore the importance of systematic mapping the long-term clearance efficacy of drug combinations for designing more-effective, resistance-proof multidrug regimes.

A naturally inspired antibiotic to target multidrug-resistant pathogens.

Gram-negative bacteria are responsible for an increasing number of deaths caused by antibiotic-resistant infections1,2. The bacterial natural product colistin is considered the last line of defence against a number of Gram-negative pathogens. The recent global spread of the plasmid-borne mobilized colistin-resistance gene mcr-1 (phosphoethanolamine transferase) threatens the usefulness of colistin3. Bacteria-derived antibiotics often appear in nature as collections of similar structures that are encoded by evolutionarily related biosynthetic gene clusters. This structural diversity is, at least in part, expected to be a response to the development of natural resistance, which often mechanistically mimics clinical resistance. Here we propose that a solution to mcr-1-mediated resistance might have evolved among naturally occurring colistin congeners. Bioinformatic analysis of sequenced bacterial genomes identified a biosynthetic gene cluster that was predicted to encode a structurally divergent colistin congener. Chemical synthesis of this structure produced macolacin, which is active against Gram-negative pathogens expressing mcr-1 and intrinsically resistant pathogens with chromosomally encoded phosphoethanolamine transferase genes. These Gram-negative bacteria include extensively drug-resistant Acinetobacter baumannii and intrinsically colistin-resistant Neisseria gonorrhoeae, which, owing to a lack of effective treatment options, are considered among the highest level threat pathogens4. In a mouse neutropenic infection model, a biphenyl analogue of macolacin proved to be effective against extensively drug-resistant A. baumannii with colistin-resistance, thus providing a naturally inspired and easily produced therapeutic lead for overcoming colistin-resistant pathogens.

Teixobactin kills bacteria by a two-pronged attack on the cell envelope.

Antibiotics that use novel mechanisms are needed to combat antimicrobial resistance1-3. Teixobactin4 represents a new class of antibiotics with a unique chemical scaffold and lack of detectable resistance. Teixobactin targets lipid II, a precursor of peptidoglycan5. Here we unravel the mechanism of teixobactin at the atomic level using a combination of solid-state NMR, microscopy, in vivo assays and molecular dynamics simulations. The unique enduracididine C-terminal headgroup of teixobactin specifically binds to the pyrophosphate-sugar moiety of lipid II, whereas the N terminus coordinates the pyrophosphate of another lipid II molecule. This configuration favours the formation of a ß-sheet of teixobactins bound to the target, creating a supramolecular fibrillar structure. Specific binding to the conserved pyrophosphate-sugar moiety accounts for the lack of resistance to teixobactin4. The supramolecular structure compromises membrane integrity. Atomic force microscopy and molecular dynamics simulations show that the supramolecular structure displaces phospholipids, thinning the membrane. The long hydrophobic tails of lipid II concentrated within the supramolecular structure apparently contribute to membrane disruption. Teixobactin hijacks lipid II to help destroy the membrane. Known membrane-acting antibiotics also damage human cells, producing undesirable side effects. Teixobactin damages only membranes that contain lipid II, which is absent in eukaryotes, elegantly resolving the toxicity problem. The two-pronged action against cell wall synthesis and cytoplasmic membrane produces a highly effective compound targeting the bacterial cell envelope. Structural knowledge of the mechanism of teixobactin will enable the rational design of improved drug candidates.

Cefepime-Taniborbactam in Complicated Urinary Tract Infection.

BACKGROUND: Carbapenem-resistant Enterobacterales species and multidrug-resistant Pseudomonas aeruginosa are global health threats. Cefepime-taniborbactam is an investigational ß-lactam and ß-lactamase inhibitor combination with activity against Enterobacterales species and P. aeruginosa expressing serine and metallo-ß-lactamases. METHODS: In this phase 3, double-blind, randomized trial, we assigned hospitalized adults with complicated urinary tract infection (UTI), including acute pyelonephritis, in a 2:1 ratio to receive intravenous cefepime-taniborbactam (2.5 g) or meropenem (1 g) every 8 hours for 7 days; this duration could be extended up to 14 days in case of bacteremia. The primary outcome was both microbiologic and clinical success (composite success) on trial days 19 to 23 in the microbiologic intention-to-treat (microITT) population (patients who had a qualifying gram-negative pathogen against which both study drugs were active). A prespecified superiority analysis of the primary outcome was performed after confirmation of noninferiority. RESULTS: Of the 661 patients who underwent randomization, 436 (66.0%) were included in the microITT population. The mean age of the patients was 56.2 years, and 38.1% were 65 years of age or older. In the microITT population, 57.8% of the patients had complicated UTI, 42.2% had acute pyelonephritis, and 13.1% had bacteremia. Composite success occurred in 207 of 293 patients (70.6%) in the cefepime-taniborbactam group and in 83 of 143 patients (58.0%) in the meropenem group. Cefepime-taniborbactam was superior to meropenem regarding the primary outcome (treatment difference, 12.6 percentage points; 95% confidence interval, 3.1 to 22.2; P = 0.009). Differences in treatment response were sustained at late follow-up (trial days 28 to 35), when cefepime-taniborbactam had higher composite success and clinical success. Adverse events occurred in 35.5% and 29.0% of patients in the cefepime-taniborbactam group and the meropenem group, respectively, with headache, diarrhea, constipation, hypertension, and nausea the most frequently reported; the frequency of serious adverse events was similar in the two groups. CONCLUSIONS: Cefepime-taniborbactam was superior to meropenem for the treatment of complicated UTI that included acute pyelonephritis, with a safety profile similar to that of meropenem. (Funded by Venatorx Pharmaceuticals and others; CERTAIN-1 ClinicalTrials.gov number, NCT03840148.).

The evolution of antimicrobial peptide resistance in Pseudomonas aeruginosa is severely constrained by random peptide mixtures.

The prevalence of antibiotic-resistant pathogens has become a major threat to public health, requiring swift initiatives for discovering new strategies to control bacterial infections. Hence, antibiotic stewardship and rapid diagnostics, but also the development, and prudent use, of novel effective antimicrobial agents are paramount. Ideally, these agents should be less likely to select for resistance in pathogens than currently available conventional antimicrobials. The usage of antimicrobial peptides (AMPs), key components of the innate immune response, and combination therapies, have been proposed as strategies to diminish the emergence of resistance. Herein, we investigated whether newly developed random antimicrobial peptide mixtures (RPMs) can significantly reduce the risk of resistance evolution in vitro to that of single sequence AMPs, using the ESKAPE pathogen Pseudomonas aeruginosa (P. aeruginosa) as a model gram-negative bacterium. Infections of this pathogen are difficult to treat due the inherent resistance to many drug classes, enhanced by the capacity to form biofilms. P. aeruginosa was experimentally evolved in the presence of AMPs or RPMs, subsequentially assessing the extent of resistance evolution and cross-resistance/collateral sensitivity between treatments. Furthermore, the fitness costs of resistance on bacterial growth were studied and whole-genome sequencing used to investigate which mutations could be candidates for causing resistant phenotypes. Lastly, changes in the pharmacodynamics of the evolved bacterial strains were examined. Our findings suggest that using RPMs bears a much lower risk of resistance evolution compared to AMPs and mostly prevents cross-resistance development to other treatments, while maintaining (or even improving) drug sensitivity. This strengthens the case for using random cocktails of AMPs in favour of single AMPs, against which resistance evolved in vitro, providing an alternative to classic antibiotics worth pursuing.

Widespread fungal-bacterial competition for magnesium lowers bacterial susceptibility to polymyxin antibiotics.

Fungi and bacteria coexist in many polymicrobial communities, yet the molecular basis of their interactions remains poorly understood. Here, we show that the fungus Candida albicans sequesters essential magnesium ions from the bacterium Pseudomonas aeruginosa. To counteract fungal Mg2+ sequestration, P. aeruginosa expresses the Mg2+ transporter MgtA when Mg2+ levels are low. Thus, loss of MgtA specifically impairs P. aeruginosa in co-culture with C. albicans, but fitness can be restored by supplementing Mg2+. Using a panel of fungi and bacteria, we show that Mg2+ sequestration is a general mechanism of fungal antagonism against gram-negative bacteria. Mg2+ limitation enhances bacterial resistance to polymyxin antibiotics like colistin, which target gram-negative bacterial membranes. Indeed, experimental evolution reveals that P. aeruginosa evolves C. albicans-dependent colistin resistance via non-canonical means; antifungal treatment renders resistant bacteria colistin-sensitive. Our work suggests that fungal-bacterial competition could profoundly impact polymicrobial infection treatment with antibiotics of last resort.