Bacterial meningitis remains difficult to treat because the blood-brain barrier restricts the entry of many antimicrobial agents into the central nervous system. The researchers’ background notes that more than 98% of small-molecule drugs and nearly all macromolecules cannot effectively penetrate the brain, potentially leaving antibiotic concentrations in cerebrospinal fluid inadequate. Conventional therapy includes third-generation cephalosporins and vancomycin, but resistance and inadequate drug exposure continue to complicate treatment. For the therapeutic experiments, the investigators selected cefoperazone, a third-generation cephalosporin used in the study against E. coli and S. aureus, because it combined antibacterial activity with intrinsically limited penetration across an intact barrier. The work therefore tested whether a designed carrier could improve delivery rather than relying on a new antibiotic molecule.

The nanoparticles were assembled through an oil-water, CTAB-templated reaction involving tetraethyl orthosilicate, producing spiky, virus-shaped particles approximately 40–50 nanometres in diameter. Their surfaces were aminated and then chemically coupled to Boc-L-alanine, whose protecting group was subsequently removed. Electron microscopy showed that this functionalization did not significantly change the particles’ morphology or size. Nitrogen-sorption analysis indicated an ordered mesoporous structure with an average pore diameter of 3.37 nanometres, while thermogravimetric analysis estimated an L-alanine grafting density of approximately 15.18%. Infrared and X-ray photoelectron spectroscopy supported the presence of alanine-derived amide and nitrogen-containing groups, and circular dichroism confirmed the intended stereochemical configuration. For comparison, the researchers prepared D-alanine-modified virus-shaped particles, conventional spherical mesoporous silica nanoparticles and an unmodified virus-shaped formulation.

After intravenous administration in healthy mice, fluorescent CVSN rapidly entered the systemic circulation and accumulated most strongly in the liver, consistent with the liver’s role in clearing silica-based nanoparticles. Substantial fluorescence was also detected in the lungs and brain. Brain signal peaked at two hours and then declined, although residual fluorescence remained at later sampling times. An independent measurement of tissue silicon by inductively coupled plasma mass spectrometry confirmed the same temporal pattern: brain silicon was significantly higher at two hours than at six or 24 hours. Brain accumulation followed the order CVSN greater than the D-alanine counterpart greater than unmodified virus-shaped particles, while D-alanine-modified particles also accumulated more than spherical particles. These comparisons indicate contributions from both particle shape and surface handedness, although they do not establish a single mechanism responsible for every difference.

Microscopy supplied evidence that CVSN moved beyond vascular attachment and entered brain parenchyma. In cortical sections, fluorescent particles appeared at vessel walls and within the surrounding tissue, with measured blood-brain barrier permeability increasing over time and reaching approximately 35% at eight hours. Bio-transmission electron microscopy detected particles within endothelial cells, astrocytes and microglia, as well as near neuronal mitochondria and in synaptic clefts. Quantified uptake in microglia was 1.539 for CVSN compared with 0.620 for unmodified virus-shaped particles, and CVSN also showed higher fluorescence in astrocytes and neurons. In mice with either E. coli or S. aureus meningitis, brain fluorescence exceeded that in healthy mice, consistent with greater barrier permeability during infection, but the experiment did not separate every possible contributor to that increase.

Cellular experiments implicated macropinocytosis and caveolae-mediated endocytosis in the uptake of CVSN by brain endothelial cells. Chlorpromazine, used to inhibit clathrin-mediated uptake, reduced intracellular fluorescence by 12.73%, whereas cytochalasin D and methyl-beta-cyclodextrin reduced it by 87.31% and 74.90%, respectively. The D-alanine control displayed a broadly similar pathway profile, while spherical particles relied more heavily on clathrin-mediated uptake. Molecular docking found favorable and broadly comparable binding energies for L- and D-alanine with proteins associated with clathrin, macropinocytosis and caveolae, although the L-alanine particles formed more hydrogen bonds in the tested comparisons. Particles bearing L- or D-cysteine and L- or D-arginine also exhibited stereochemistry-dependent differences in endothelial uptake and brain accumulation. Together, these findings support a role for surface handedness interacting with side-chain chemistry rather than proving one universal molecular interaction.

Alanine density also showed a non-linear relationship with biological performance. Formulations prepared with half and twice the standard L-alanine feed entered the circulation and reached the brain, but neither surpassed the standard formulation in brain accumulation or endothelial uptake. The standard preparation generated stronger cellular fluorescence than unmodified particles and both the lower- and higher-ligand versions, indicating an optimal ligand presentation among the formulations tested. CVSN retained its principal structural signals during storage and serum incubation and degraded gradually in phosphate-buffered saline and artificial cerebrospinal fluid. A major limitation is that the carrier had no brain- or disease-responsive release gate. Release was assessed in laboratory media, and detectable peripheral distribution means that premature release and off-target antibiotic exposure cannot be excluded.

CVSN accommodated cefoperazone, norfloxacin, levofloxacin and imipenem, which differ in chemical class, molecular weight, solubility, chirality, antibacterial spectrum and inherent blood-brain barrier permeability. Drug-loading capacities ranged from 25.67% for imipenem to 41.29% for cefoperazone, with encapsulation efficiencies of 51.33% to 82.57%. Loading cefoperazone filled the mesopores, reduced measurable surface area and apparently converted the crystalline drug into an amorphous state. In a bEnd.3 endothelial monolayer whose resistance stabilized above 150 millivolts, free cefoperazone crossed with an efficiency of 0.65%, compared with 9.18% for CVSN@CPZ after two hours. In mice, CVSN-loaded cefoperazone produced higher brain concentrations than free drug at one, two and six hours. Loading also significantly increased brain-to-plasma and cerebrospinal-fluid-to-plasma ratios for all four tested antibiotics, demonstrating broader applicability across the selected drugs under the tested doses and sampling intervals.

In laboratory assays, CVSN@CPZ lowered the minimum inhibitory concentration of cefoperazone from 16 to 8 micrograms per millilitre for E. coli and from 512 to 256 micrograms per millilitre for S. aureus. Unloaded CVSN had no obvious antibacterial effect. Electron microscopy indicated bacterial cell-wall disruption, while colony counting, crystal-violet staining and propidium-iodide staining supported stronger inhibition of growth, biofilm formation and bacterial death with the loaded carrier. In experimentally infected mice receiving daily intravenous treatment for five days, the reported inhibition rate against E. coli increased from 96.69% with free cefoperazone to 97.64% with CVSN@CPZ. The improvement against S. aureus was larger, increasing from 79.66% with free cefoperazone to 94.75% with the loaded carrier.

Survival and neurological outcomes also differed among the mouse treatment groups. In the E. coli model, survival was 100% with CVSN@CPZ, compared with 66.67% for free cefoperazone and 50% in untreated animals. In the S. aureus model, survival reached 83.33% with CVSN@CPZ, while both free-drug and untreated groups had 50% survival. CVSN@CPZ reduced brain TNF-alpha, IL-1beta and IL-6 concentrations, preserved the tight-junction protein ZO-1 and reduced markers of neutrophil and activated myeloid-cell infiltration. Tissues treated with the loaded carrier showed lower CD86 and higher CD206 expression, consistent with less pro-inflammatory and more reparative macrophage phenotypes, along with reduced caspase-3 staining and suppression of the TLR4/TLR9–MyD88–NF-kappaB signaling axis. Histology showed less meningeal inflammation and tissue disruption. These measurements were obtained in experimentally infected mice and do not establish equivalent safety, dosing or efficacy in patients. Overall, the study links chiral surface modification and virus-like architecture to improved preclinical antibiotic delivery, while clinical translation, large-animal validation, long-term toxicity and spatially controlled release remain unresolved.

Subject of Research: Preclinical evaluation of a chiral virus-like nanoplatform for brain delivery of antibiotics in bacterial meningitis.

Article Title: A Chiral Virus‐Like Nanoplatform That Broadly Mediates Antibiotic Delivery to the Brain for Bacterial Meningitis Therapy

Article References: Xu, Q., Zheng, X., Tong, H., Liu, Y., Fan, X., Xu, M., Zhang, Z., Zhang, L., Liu, M., Du, T., Li, H., & Liu, G. (2026). A Chiral Virus‐Like Nanoplatform That Broadly Mediates Antibiotic Delivery to the Brain for Bacterial Meningitis Therapy. Advanced Science, Article e78317. https://doi.org/10.1002/advs.78317

Image Credits: AI Generated

DOI: 10.1002/advs.78317

Keywords: bacterial meningitis, blood-brain barrier, antibiotic delivery, chiral nanoparticles, mesoporous silica, cefoperazone, virus-like nanoplatform, mouse models, neuroinflammation, Chiral, Virus, Like