Exploring lipid nanoparticle design spaces using self-regulating microfluidic machines and multiplexed in vivo biodistribution

Delivering therapeutic mRNA relies on lipid nanoparticles (LNPs). Finding optimal process parameters for new lipid combinations in LNP formulations remains a challenge. In our work, we used an automated, self-regulating microfluidic platform that actively changes process parameters to tune LNP formulations for preset, desired quality standards. We tested new LNPs by swapping poly(ethylene glycol) (PEG) lipopolymers for alternatives based on poly(2-methyl-2-oxazoline) (PMeOx) and poly(2-ethyl-2-oxazoline) (PEtOx). For each lipopolymer variant, the platform independently identified optimal production conditions in four or fewer iterative cycles, yielding particles of the preset size and high mRNA encapsulation. Small-angle X-ray scattering revealed that smaller LNPs modified with PEtOx had more structural surface variety and higher mRNA loading efficiency. When multiplexing these formulations in mice, the PEtOx-containing LNPs accumulated more in bone marrow compared to those with PEG, indicating trends that the chemistry of the lipopolymer affects the biodistribution of the resulting LNPs. By combining automated formulation and in vivo multiplexed testing, our approach provides a practical way to rapidly plan, formulate, and evaluate large pharmaceutical design spaces, to select excipients and process parameters yielding optimal biological performance of LNPs.