A molecular architectural design that promises potent antimicrobial activity against multidrug-resistant pathogens

The study addresses the urgent need for new antimicrobial strategies in the face of rising drug resistance. Traditional drug design evolved from modifying natural products to rational design incorporating medicinal, molecular, and quantum chemistry. Advances in biophysics now enable understanding of modes of action of functional units within drugs. The research question is whether designing synthetic antibacterial complexes with specialized architectures, built from melittin (a natural antimicrobial peptide) and PEG (a clinically used polymer), can enhance antibacterial potency while reducing toxicity. The purpose is to create architecture-modulated agents that exploit membrane-targeting mechanisms to outperform existing antibiotics and overcome multidrug resistance. The importance lies in offering a new, biophysics-guided molecular architectural approach to produce highly efficient, lower-toxicity antimicrobials.

Prior work integrating functional units has shown promise: a β-hairpin macrocycle targeting LptD of Gram-negative bacteria was integrated with the colistin peptide macrocycle that binds lipid A of LPS, yielding an agent active against colistin-resistant pathogens and advancing to preclinical toxicology. Self-assembled protein/peptide-based hybrid nanostructures (e.g., collagen–gold hydrogels, Fmoc amino acid–Ag biomaterial hydrogels, Ce6–doxorubicin nanoparticles, and glutaraldehyde-assisted dipeptide–Ce6–heparin particles) demonstrate combined antimicrobial, antitumor, and tissue culture applications. While combining existing units can improve performance, the paper argues that stronger biofunctions can be achieved by designing specialized architectures inspired by natural protein scaffolds, motivating the PEG–melittin constructs evaluated here.

Synthesis of PEG–melittin complexes: PEGs bearing reactive groups were covalently conjugated to melittin via coupling between PEG active sites (maleimide-derived) and peptide amino groups at a 2.5:1 molar ratio (active site:peptide). Conjugates (e.g., PEG12k–n*Mel series) were dialyzed (MWCO 3500), concentrated, and characterized by dynamic light scattering (DLS), SDS–PAGE, UV–Vis, and FT-IR. The nomenclature denotes PEG molecular weight and the number of melittin peptides per PEG (e.g., PEG12k–1*Mel). Characterization included SDS–PAGE to verify conjugation; DLS to assess size and concentration-dependent self-assembly. GUV leakage assay: Giant unilamellar vesicles composed of DOPC and/or DOPG with 1 mol% Rh-PE were prepared by electroformation. After addition of agents, calcein influx into GUV interiors was monitored via confocal microscopy over time. Fluorescence intensity inside GUVs was normalized to the surrounding medium and plotted versus time to quantify membrane permeabilization dynamics. Antimicrobial testing: Minimal inhibitory concentrations (MICs) were determined using standardized protocols against Gram-negative and Gram-positive strains, including Escherichia coli CGMCC 1.12883, Staphylococcus aureus CGMCC 1.10755, and multidrug-resistant clinical or reference strains. Time–kill assays quantified bactericidal dynamics over 0.3, 6, and 16 h exposures, comparing PEG12k–1*Mel with pristine melittin. Acute toxicity in mice: C57BL/6 female mice (20 ± 2 g) received intraperitoneal injections of melittin (5 mg kg−1), PEG12k–1*Mel (5 mg kg−1), or PBS. After 24 h, serum ALT and creatinine were measured; liver and kidney tissues were collected for H&E staining and histopathology. Molecular dynamics simulations: Coarse-grained (CG) MD with Martini force field (v2.2) implemented in GROMACS 5.1.4. DLPG lipids were mapped using a 4:1 scheme; PEG was modeled with Martini SPo beads. Periodic boundary conditions were applied in all directions to study membrane interactions and pore/channel formation behavior. Statistics: Experiments and simulations were repeated three to five times per condition. Data are reported as mean ± SD. ANOVA (OriginPro 9.0) assessed differences, with comparisons to controls unless stated otherwise. Additional details are in the Supplementary Information.

- Architectural design: A series of PEG–melittin complexes with varying PEG structures and peptide valencies were synthesized. The flexibly linear PEG12k–1*Mel (one melittin conjugated to ~12 kDa PEG) exhibited the best performance. - Self-assembly: PEG12k–1*Mel formed concentration-dependent aggregates by DLS, increasing from ~500 nm at 0.1 mg mL−1 to ~800 nm at 5.0 mg mL−1, unlike pristine melittin (dispersed random coils). - Potent antibacterial activity: Against E. coli and S. aureus, PEG12k–1*Mel caused severe membrane disruption and content leakage by SEM after 16 h at 64 μg mL−1 (peptide basis). Mean MICs for PEG12k–1*Mel were 6 μg mL−1 (E. coli) and 17 μg mL−1 (S. aureus), versus 32 and 45 μg mL−1, respectively, for pristine melittin. Both agents outperformed polymyxin B (E. coli) and fusidic acid (S. aureus) in these assays. - Activity against MDR pathogens: Reported MIC range across multidrug-resistant pathogens was 2–32 μg mL−1. Representative MICs (μg mL−1): A. baumannii ATCC19606: PEG12k–1*Mel 4, Mel 128, Ceftazidime 8, Vancomycin 64, Ampicillin >256; P. aeruginosa ATCC27853: PEG12k–1*Mel 64, Mel 128, Vancomycin 64, Ampicillin >256; A. baumannii clinical isolate 1: PEG12k–1*Mel 32, Mel >128, Ceftazidime 256, Meropenem 64, Vancomycin 128, Ampicillin >256; A. baumannii clinical isolate 3: PEG12k–1*Mel 2, Mel 128, Ceftazidime 256, Meropenem 64, Vancomycin >256, Ampicillin >256; MRSA: PEG12k–1*Mel 16, Mel >128. - Time–kill dynamics: Bacterial viability after exposure to PEG12k–1*Mel at 0.3, 6, and 16 h was consistently lower than with pristine melittin, indicating faster and stronger bactericidal action. - Mode of action: Biophysical assays and simulations indicate lipid-specific membrane recognition and an accelerated “channel” (pore) formation effect by the PEG–melittin complex, enhancing membrane perforation relative to melittin alone. - Safety profile: In vivo, PEG12k–1*Mel reduced cytotoxicity by 68% and acute toxicity by 57% relative to melittin. Mouse serum biomarkers (ALT, creatinine) and H&E histology supported improved hepatic/renal safety at 5 mg kg−1 dosing. - Overall improvement: The best architecture (PEG12k–1*Mel) delivered up to 500% improvement in antimicrobial efficiency compared with pristine melittin, while lowering toxicity.

The findings show that rational architectural design of a natural AMP with a clinically used polymer can markedly modulate antimicrobial potency and safety. Conjugating a single melittin to a long-chain PEG (12 kDa) produced a flexibly linear construct that self-assembles and preferentially interacts with bacterial membrane lipids (DOPG-rich), facilitating faster, more efficient pore formation. This architecture increases bacterial membrane disruption (SEM evidence) and accelerates intracellular leakage (GUV assays), resulting in lower MICs and superior time–kill kinetics compared with melittin. Notably, efficacy extended to multidrug-resistant Gram-negative and Gram-positive pathogens, with MICs as low as 2–32 μg mL−1 and substantial improvements over melittin and some conventional antibiotics in the tested panel. The PEGylation also mitigated systemic toxicity, likely by modulating peptide membrane selectivity and pharmacodynamics, as reflected by decreased ALT and creatinine and reduced histopathological damage in mice. Coarse-grained MD simulations support a lipid-specific recognition and enhanced channel formation mechanism for the PEG–melittin complex, offering a molecular basis for the observed potency and selectivity. Together, the results address the research question by demonstrating that specialized molecular architectures can enhance antimicrobial action while reducing toxicity, highlighting an architectural, biophysics-guided avenue for AMP-based drug development against multidrug-resistant pathogens.

This work introduces a biophysical architectural design strategy for antimicrobial agents, exemplified by PEG12k–1*Mel, which delivers up to 5-fold higher antimicrobial efficiency than melittin, potent activity against multidrug-resistant bacteria (MIC 2–32 μg mL−1), and significantly reduced in vivo cytotoxicity and acute toxicity. Mechanistically, the designed architecture enhances lipid-specific membrane recognition and accelerates pore formation, leading to superior bactericidal performance. These results demonstrate the potential of minimal-component architectural constructions (natural AMP + clinically used polymer) to yield high-efficacy, lower-toxicity antimicrobials. Future work can expand architectural variations (PEG length/branching, valency), broaden pathogen coverage, elucidate detailed molecular mechanisms with higher-resolution simulations and experiments, and assess in vivo efficacy and safety in infection models to progress toward clinical translation.