Autonomous Chemotactic Nanoparticles for Enhanced Blood–brain Barrier Navigation via Dynamic Gradient Sensing and Adaptive Surface Remodelling

Abstract Introduction: The blood–brain barrier (BBB) restricts the entry of most therapeutic molecules into the central nervous system (CNS), limiting treatment options for neurodegenerative and neuroinflammatory diseases. Conventional nanoparticle (NP) approaches, whether passive (size and charge optimisation) or active (ligand-receptor targeting), lack adaptability to dynamic BBB environments. Inspired by bacterial chemotaxis, we propose autonomous chemotactic NPs (Chemo-NPs) capable of sensing endogenous gradients such as glutamate and ATP and dynamically remodelling their surface to optimise BBB penetration. Methods: We engineered lipid-polymer hybrid NPs incorporating synthetic GPCR-based chemoreceptors, an enzymatic phosphorylation cascade for surface charge modulation and PEG-PBAE brushes enabling dual transport modes (passive diffusion and adsorptive-mediated transcytosis). Microfluidic devices generated stable chemical gradients, while high-speed three-dimensional microscopy tracked particle trajectories. A transwell-based in vitro BBB model was used to assess penetration. Comparative groups included static PEG-NPs, receptor-only NPs and the full chemotactic system. Computational modelling using a biased random walk framework complemented experimental analysis. Results: Chemo-NPs demonstrated concentration-dependent directional movement, achieving a chemotactic index of 0.72 ± 0.08 in glutamate gradients, compared to 0.41 ± 0.05 for receptor-only NPs and 0.11 ± 0.03 for PEG-NPs ( P < 0.001). Surface charge modulation occurred dynamically, with zeta potential switching from −5.2 to +18.3 mV in response to ATP gradients. In the BBB model, Chemo-NPs achieved 6.8% ± 0.9% translocation – 4.5-fold higher than PEG-NPs and 2.7-fold higher than receptor-only NPs. Performance was enhanced under the acidic conditions (pH 6.8), mimicking inflamed BBB environments. Computational predictions closely matched the experimental results ( R 2 > 0.84). Conclusion: This work demonstrates the feasibility of autonomous Chemo-NPs for dynamic BBB navigation. By integrating gradient sensing with adaptive surface remodelling, Chemo-NPs outperform static formulations, offering a versatile platform for CNS drug delivery. Future efforts should address in vivo stability, minimising off-target accumulation and translational scalability. The modular design further suggests potential applications in inflammatory and tumour microenvironments, marking a step towards programmable nanomedicines with autonomous decision-making capabilities.