Yuchen Fan

Limited chromatographic resolution of large nucleic acids remains a major bottleneck in the development of high-throughput analytical platforms for mRNA therapeutics. Here, we report a pronounced on-off elution behavior of oligonucleotides and mRNA under strong ion-exchange (IEX) conditions, using a normalized linear ionic strength model. This discovery fundamentally redefines solvent-strength scaling in IEX and enables rapid separations of oligonucleotides and mRNA on ultra-short columns. Leveraging this on-off elution behavior, two complementary high-throughput IEX methods were developed using 5 cm columns under a unified platform (pH 8, 75 °C, lithium perchlorate as the mobile-phase additive). A 6.5-minute gradient achieved baseline separation and quantification of non-poly(A)- and poly(A)-mRNA species for both eGFP and Fluc mRNA, representing the first reported baseline chromatographic separation of intact non-poly(A)- and poly(A)-mRNA. To resolve poly(A) length variants within the oligonucleotide sequence range, a non-linear logarithmic gradient achieved separation of 20-120 nt poly(A) ladders within 4 minutes, enabling high-throughput poly(A) tail length determination. Following one-step RNase T1 digestion, poly(A) tail lengths of 117 nt and 118 nt were determined for eGFP and Fluc mRNA, respectively, in agreement with theoretical values. Collectively, this work establishes on-off elution behavior as a defining retention regime in IEX and demonstrates it as a powerful strategy for high-throughput characterization of mRNA poly(A) content and tail length, with strong potential to support process development and quality control of mRNA therapeutics.

Hydrophobic interaction chromatography (HIC) is a current gold standard for determining the drug-to-antibody ratio (DAR) and drug load distribution (DLD) in antibody-drug conjugates (ADCs). In this study, we developed several HIC methods using ammonium tartrate for comprehensive ADC characterization and bioconjugation reaction monitoring. We first demonstrated that ammonium tartrate delivers a separation performance comparable to conventional ammonium sulfate while offering superior mass spectrometry compatibility. We then established optimized platform HIC methods for analyzing both interchain and engineered cysteine-conjugated ADCs. Furthermore, we developed a rapid 10 min, multiattribute HIC method as a process analytical technology (PAT) for real-time bioconjugation reaction monitoring. This multiattribute HIC PAT method was used to simultaneously track DAR, DLD, and drug-linker (DL) concentration within complex reaction mixtures. Through advanced PAT tools, we gained critical mechanistic insights into the kinetics of ADC conjugation. For example, we discovered that the DL dissolution rate could limit reaction progression in slurry reactions, an insight not possible through conventional technologies such as ultraviolet spectroscopy. These findings enabled the identification of critical process parameters, informing targeted process development that ultimately reduced reaction times by 90% with a designed average DAR and lower undesired overconjugation.

Lipid nanoparticles (LNPs) are a powerful drug delivery platform advancing vaccines and gene therapies. While their efficacy and safety has been found to be closely linked to innate immune activation, current in vitro models are unable to predict immune responses reliably. Conventional models, such as PBMCs, are limited by donor variability and inconsistent sensitivity. To address this, we developed a cytokine profiling platform using induced pluripotent stem cell (iPSC)-derived monocytes (iMonocytes), a physiologically relevant innate immune cell type that plays a key role in immune surveillance and inflammation. iPSCs provide a renewable, uniform monocyte source for consistent, high-sensitivity LNP screening. When tested with LNPs of graded immunostimulatory potency, iMonocytes showed improved reproducibility and strong correlation with in vivo cytokine responses. This platform enables evaluation of cargo- and dose-dependent effects, providing a robust and scalable tool for preclinical assessment and rational design of LNP therapeutics.

In recent years, the use of lipid nanoparticles (LNPs) for delivery of messenger RNA (mRNA)–based therapies has gained substantial attention in the field of drug development. In such an application, multiple LNP attributes have to be carefully characterized to ensure product safety and quality, whereas accurate and efficient characterization of these complex mRNA-LNP formulations remains a challenging endeavor. Here, we present the development and application of an online separation and characterization platform designed for the isolation and in-depth analysis of mRNAs and mRNA-loaded LNPs. Our asymmetrical flow field-flow fractionation with a multi-detector (MD-AF4) method has demonstrated exceptional resolution between mRNA-LNPs and mRNAs, delivering excellent recoveries (over 70%) for both analytes and exceptional repeatability. Notably, this platform allows for comprehensive and multi-attribute LNP characterization, including online particle sizing, morphology characterization, and determination of encapsulation efficiency, all within a single injection. Furthermore, real-time online sizing by synchronizing multi-angle light scattering (MALS) and dynamic light scattering (DLS) presented higher resolution over traditional batch-mode DLS, particularly in differentiating heterogeneous samples with a low abundance of large-sized particles. Additionally, our method proves to be a valuable tool for monitoring LNP stability under varying stress conditions. Our work introduces a robust and versatile analytical platform using MD-AF4 that not only efficiently provides multi-attribute characterizations of mRNA-LNPs but also holds promise in advancing studies related to formulation screening, quality control, and stability assessment in the evolving field of nanoparticle delivery systems for mRNAs. Graphical Abstract