The urgency behind the search is hard to overstate. MRSA and the other so-called ESKAPE pathogens, including Enterococcus faecium, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa and drug-resistant Enterobacter species, sit at the center of a worsening global crisis. MRSA strains carry the mecA gene, which encodes a penicillin-binding protein with low affinity for beta-lactam antibiotics, allowing the bacteria to keep building their cell walls in the presence of drugs that once destroyed them. With multidrug-resistant strains spreading faster than new antibiotics are being developed, scientists have turned increasingly to underexplored ecological niches, and endophytic fungi, the microorganisms that live within plant tissues without harming their hosts, have proven to be especially rich sources of structurally diverse bioactive metabolites.

What makes the new study particularly striking is how little was known about its star organism. Only about 25 publications on the genus Neocucurbitaria existed prior to this work, most focusing on fungal ecology, and just three had examined secondary metabolites. Those earlier studies, conducted on the marine fungus Neocucurbitaria unguis-hominis, yielded 17 diterpenoids, most of which showed little antibacterial or cytotoxic activity. The Groningen team, led by Kristina Haslinger and Jan Maarten van Dijl with first author Xiao Li, took a different route: they had previously built an in-house library of endophytic fungi from Vinca minor, and when they screened crude ethyl acetate extracts for antibacterial activity, isolate VM-36 consistently showed one of the strongest inhibitory effects against the indicator bacterium Bacillus subtilis.

To understand what the fungus was capable of, the researchers sequenced its entire genome using long-read Oxford Nanopore technology, generating 870,296 raw reads that assembled into a 32.8-megabase genome across 19 contigs, with a completeness score of 97.2 percent. The assembly predicted 11,817 protein-coding genes, along with 120 transfer RNA genes and 29 ribosomal RNA genes. Phylogenetic analysis using multiple barcode loci placed the fungus close to Neocucurbitaria salicis-albae, a species previously isolated from a willow twig in Germany. More importantly, genome mining with the antiSMASH pipeline revealed 34 putative biosynthetic gene clusters, a number comparable to the average of 36 found in related Cucurbitariaceae fungi, spanning polyketide synthases, non-ribosomal peptide synthetases, ribosomally synthesized peptide clusters, terpene clusters and hybrids.

Four of those gene clusters stood out because they appear to be unique to VM-36, with similarities to clusters known to produce the antifungal agent griseofulvin, the antibacterial compounds usnic acid and phomasetin, and the anticancer agent hypothemycin. The researchers also found evidence of a large carbohydrate-active enzyme arsenal, with 571 related genes, hinting at the fungus’s ecological role in decomposing plant biomass inside its host. On the metabolomic side, untargeted liquid chromatography-tandem mass spectrometry and molecular networking on the GNPS platform produced a network of 1,521 nodes and 1,896 edges, from which 19 compounds were annotated by library matching and 7 more were putatively assigned. Among the most abundant constituents were analogues of CJ-21058, a known fungal SecA inhibitor, belonging to the decalin-containing tetramic acid family.

The team then scaled up fermentation, growing the fungus in 13 liters of dextrose peptone yeast medium for 30 days and purifying the dominant compound, designated compound 1, to 114 milligrams. Using nuclear magnetic resonance spectroscopy, mass spectrometry, Marfey’s analysis of the acid hydrolysate, and electronic circular dichroism supported by time-dependent density functional theory calculations, they established that compound 1 is a decalin-containing tetramic acid with the molecular formula C23H33NO4, closely related to the known compounds CJ-21058 and hyalodendrin A. The serine-derived tetramic acid moiety was determined to have the L-configuration, and the absolute configuration was tentatively assigned as (2R, 3S, 6R, 8S, 11S, 5’S). Intriguingly, the compound appears to exist as a mixture of two tautomeric forms differing in where the keto and enol groups sit within the tetramic acid ring.

The antibacterial results are what elevate the compound to headline status. Against a panel of ESKAPE(E) pathogens, compound 1 showed strong activity against Gram-positive bacteria, including Enterococcus faecalis and multiple Staphylococcus aureus strains, but was entirely inactive against Gram-negative species such as Escherichia coli, Klebsiella pneumoniae and Pseudomonas aeruginosa, all with minimum inhibitory concentrations above 64 micrograms per milliliter. Against MRSA and methicillin-sensitive S. aureus alike, the compound achieved MIC values of 0.5 to 1 microgram per milliliter, matching or approaching the clinical antibiotics vancomycin and daptomycin used as positive controls. With minimum bactericidal concentrations of 2 to 4 micrograms per milliliter and MBC-to-MIC ratios at or below 4, compound 1 is clearly bactericidal, not merely bacteriostatic.

Kill kinetics reinforced that picture. When exponentially growing cultures of S. aureus strains USA300, D15, D17 and ATCC 29213 were exposed to compound 1 at 4 and 8 micrograms per milliliter, viable counts dropped by up to three orders of magnitude, sometimes to zero detectable colonies, outperforming both vancomycin and daptomycin at those concentrations. The optical density of the cultures declined after treatment, suggesting the compound may induce bacterial lysis, though the authors caution that membrane permeability assays and leakage experiments will be needed to confirm a bacteriolytic mechanism. In checkerboard combination assays, no antagonism was observed across 35 drug combinations, and compound 1 combined additively with daptomycin against the clinical MRSA isolate D17, with a fractional inhibitory concentration index of 0.625, a finding that suggests the molecule could complement existing therapies without undermining them.

Safety data, however, temper the excitement. Compound 1 caused 50 percent hemolysis of human red blood cells at 34.54 micrograms per milliliter, roughly four times its bactericidal concentration, indicating moderate hemolytic activity at effective doses. More concerning, the compound inhibited the metabolic activity of human bronchial epithelial cells with an IC50 of 2.39 micrograms per milliliter, well below its antibacterial MBC, whereas vancomycin and daptomycin showed IC50 values of 111.25 and above 200 micrograms per milliliter, respectively. That narrow selectivity window means compound 1 is far from ready for systemic use, and the authors emphasize that structural optimization, along with studies of stability, solubility, pharmacokinetics, resistance development and in vivo efficacy, will be essential before it can be considered a genuine drug lead.

Even so, the study reshapes expectations for an entire fungal genus. By connecting a candidate hybrid polyketide synthase-non-ribosomal peptide synthetase gene cluster to compound 1 production, noting that the core enzyme NeoA shares 72 percent amino acid identity with the phomasetin biosynthesis enzyme Phm1, and proposing a Diels-Alderase step catalyzed by NeoC, the researchers have laid a genomic framework for genome-guided discovery of related antibiotics. They also propose a possible biosynthetic origin for a cyclic siderophore detected in the metabolomic network. For a field racing to find new scaffolds against drug-resistant bacteria, the message is clear: some of the most promising chemical arsenals may be hidden not in exotic deep-sea sediments but in the quiet symbiosis between a common garden plant and the fungus living inside its leaves.

Subject of Research: Antibacterial metabolites from the endophytic fungus Neocucurbitaria sp. VM-36 active against MRSA

Article Title: Genomics and metabolomics analyses reveal the endophytic fungus Neocucurbitaria sp. VM-36 as a potential source of bioactive metabolites against methicillin-resistant Staphylococcus aureus

Article References: Li, X., Flores-Vallejo, R. D. C., He, T., van Dijl, J. M., & Haslinger, K. (2026). Genomics and metabolomics analyses reveal the endophytic fungus Neocucurbitaria sp. VM-36 as a potential source of bioactive metabolites against methicillin-resistant Staphylococcus aureus. Frontiers in Microbiology, 17, Article 1929829. https://doi.org/10.3389/fmicb.2026.1929829

Image Credits: AI Generated

DOI: 10.3389/fmicb.2026.1929829

Keywords: MRSA, antibiotic resistance, endophytic fungi, Neocucurbitaria, Vinca minor, decalin-containing tetramic acids, natural products, genome mining, metabolomics, Staphylococcus aureus, biosynthetic gene clusters, daptomycin

The nanoparticles at the heart of the study are built from a simple mixture of phytantriol and a phospholipid called DPPS. By adjusting the salt concentration of the surrounding solution, the researchers could drive the same lipid formulation through a sequence of distinct internal structures. In pure water, the lipids assembled into lamellar vesicles, essentially flat bilayer spheres. Dilution into 0.1-times phosphate-buffered saline transformed them into so-called P-cubosomes, particles whose interiors form a bicontinuous cubic phase with the Im3m symmetry. A further step into full-strength PBS produced D-cubosomes with the Pn3m diamond-type cubic phase. Small-angle X-ray scattering and cryogenic electron microscopy confirmed each transition, and the particles kept their assigned structures even after being transferred into the biological assay medium used for all subsequent experiments.

What makes this platform powerful is that the three particle types are nearly identical in every measurable respect except curvature. Dynamic light scattering showed hydrodynamic diameters of roughly 125 to 142 nanometers, and all three displayed a near-neutral zeta potential of about minus 3 millivolts in the assay medium. Co-administration with the antibiotic daptomycin did not alter these colloidal properties. Yet when the researchers applied the Gauss-Bonnet theorem to the X-ray-derived lattice parameters, they calculated Gaussian curvatures of zero for the vesicles, about minus 0.07 per square nanometer for P-cubosomes, and minus 0.10 per square nanometer for D-cubosomes. The cubic particles expose saddle-shaped openings where internal water channels meet the particle surface, while the vesicle surface is topologically flat. Increasing salt screens electrostatic repulsion between lipid headgroups, reduces the interfacial area per molecule, and pushes the system toward progressively more negative curvature.

With this curvature hierarchy established, the team tested how the blank, drug-free particles interacted with six clinical MRSA strains. Confocal microscopy showed all three formulations gathering near the bacterial membrane, but quantitative assays revealed a clear trend. In the NPN uptake assay, which measures penetration into the hydrophobic membrane interior, D-cubosomes achieved the highest uptake factor of 9.5 at 64 micrograms per milliliter, closely followed by P-cubosomes at 9.4, while vesicles lagged at 8.1. D-cubosomes also triggered the strongest oxidative stress in bacterial cells. The particles alone, however, could not kill the bacteria; their role emerged only in combination with daptomycin, a lipopeptide antibiotic that disrupts Gram-positive membranes by binding to negatively charged lipids.

The combination experiments produced striking results. Adding subinhibitory daptomycin to vesicle suspensions reduced the minimum inhibitory concentration of the drug eight-fold compared with nanoparticle monotherapy. With P-cubosomes the reduction reached 32-fold, and with D-cubosomes it climbed to 64-fold. Similar trends held across all six MRSA strains tested. Notably, the fractional inhibitory concentration indices remained within the conventional indifferent range, meaning the effect is best described as curvature-dependent potentiation rather than formal pharmacodynamic synergy. Control experiments confirmed that ionic strength alone did not explain the enhancement, since daptomycin activity was unchanged in buffers of different salinity without nanoparticles.

Fluorescence live-dead staining quantified the bactericidal consequences. Daptomycin alone left approximately 81 percent of cells viable, comparable to untreated controls. Combined with vesicles, viability fell to about 15 percent; with P-cubosomes, to roughly 7 percent; and with D-cubosomes, to below 2 percent. Electron microscopy added visual weight to these numbers. Untreated MRSA cells showed smooth, uniform surfaces, and neither daptomycin alone nor any nanoparticle alone caused visible damage. The daptomycin-D-cubosome combination, however, produced extensive membrane disruption, with distinctive bread-crumb-like protrusions covering the cell surface, while the vesicle combination left morphology essentially intact and P-cubosomes caused only moderate blebbing. Transmission electron microscopy further showed D-cubosomes in the closest contact with bacterial envelopes, with visible cell lysis in the combination treatment.

To probe the mechanism at molecular resolution, the researchers turned to neutron reflectometry. They first confirmed with fluorescent peptidoglycan labeling that the nanoparticles were not primarily accumulating in the cell wall, justifying a simplified model membrane that recreates the S. aureus cytoplasmic membrane from phosphatidylglycerol, cardiolipin, and lysyl-phosphatidylglycerol in the same 69:12:19 molar ratio found in the bacterium. Using three isotopic contrasts, the team measured how each formulation remodeled this bilayer. Daptomycin with vesicles removed only 9.5 percent of the lipid volume fraction. The P-cubosome combination extracted 17.8 percent, consistent with partial lipid loss. The D-cubosome combination was by far the most destructive, stripping 36 percent of the lipid and allowing water to penetrate the bilayer, and a layer associated with D-cubosomes, roughly 33.5 angstroms thick, was detected deposited on the membrane surface.

The physical explanation draws on well-established membrane biophysics. According to Helfrich’s curvature-elastic energy model, membranes store elastic energy that depends on how far their shape deviates from a spontaneous curvature, and Gaussian curvature contributes directly to this energy landscape. Negative Gaussian curvature, the saddle-like geometry characteristic of the cubic phases, lowers the energetic barrier for membrane remodeling processes such as fusion and pore formation. The MRSA membrane, dominated by lysyl-phosphatidylglycerol, naturally favors positive curvature, so highly negatively curved nanoparticles create a curvature mismatch that generates localized stress, weakens lipid packing, and increases permeability. Daptomycin at subinhibitory levels presumably sensitizes the membrane, while the retained internal curvature of the nanoparticles governs how much remodeling follows.

Safety and in vivo data strengthened the case. Human embryonic kidney cells maintained at least 80 percent metabolic activity after 24 hours of exposure to any formulation alone or combined with daptomycin. In a mouse bacteremia model, intravenous infection with S. aureus followed by intraperitoneal treatment showed a clear curvature-dependent hierarchy of bacterial clearance. Free daptomycin at 10 milligrams per kilogram, chosen to represent a suboptimal therapeutic condition, did not significantly reduce organ bacterial burdens compared with saline. Blank nanoparticles alone produced a modest two log10 reduction. The daptomycin-vesicle combination roughly doubled that effect, the P-cubosome combination lowered counts to between 10,000 and 100,000 colony-forming units per gram, and the D-cubosome combination pushed bacterial counts in the kidney, liver, and spleen close to the detection limit of 100 CFU per gram, a five to six log10 improvement over the free drug.

The authors are careful to frame the animal work as a proof of concept rather than therapeutic validation. Serum proteins, lipoproteins, and endogenous lipases could remodel the nanoparticles before they reach infected tissues, and the study did not include pharmacokinetic analysis, long-term toxicity assessment, or survival endpoints. Future development will require plasma stability studies, quantitative measurement of any daptomycin association with the lipid phases, biodistribution data, dose optimization, and immunogenicity evaluation. Even so, the central message stands out clearly: by isolating curvature as the only structural variable in a matched nanoparticle platform, the study provides direct experimental evidence that nanoscale geometry, tunable through something as simple as salt concentration, can be programmed to amplify antibiotic activity against one of the world’s most dangerous drug-resistant pathogens.

Subject of Research: Curvature-modulated lyotropic liquid crystalline nanoparticles as antibacterial potentiators against MRSA

Article Title: Enhanced Antibacterial Properties of Lyotropic Liquid Crystalline Nanoparticles via Curvature Modulation

Article References: Lai, X., Wang, S., Ding, C., Kostoulias, X., Brun, A. P. L., Hsu, H.-Y., Jiang, J.-H., Wang, Y., Strugnell, R. A., Peleg, A. Y., & Shen, H.-H. (2026). Enhanced Antibacterial Properties of Lyotropic Liquid Crystalline Nanoparticles via Curvature Modulation. Advanced Science, 13(55), Article e76516. https://doi.org/10.1002/advs.76516

Image Credits: AI Generated

DOI: 10.1002/advs.76516

Keywords: antimicrobial resistance, MRSA, lyotropic liquid crystalline nanoparticles, cubosomes, Gaussian curvature, daptomycin, membrane disruption, neutron reflectometry, nanomedicine, Staphylococcus aureus, lipid nanoparticles, antibiotic potentiation