Machine Learning Reveals Amine Type in Polymer Micelles Determines mRNA Binding, In Vitro, and In Vivo Performance for Lung-Selective Delivery DOI Creative Commons
Sidharth Panda, Ella J. Eaton,

Praveen Muralikrishnan

et al.

JACS Au, Journal Year: 2025, Volume and Issue: unknown

Published: April 14, 2025

Cationic micelles, composed of amphiphilic block copolymers with polycationic coronas, offer a customizable platform for mRNA delivery. Here, we present library 30 cationic micelle nanoparticles (MNPs) formulated from diblock reactive poly(pentafluorophenol acrylate) backbones modified diverse amines. This systematically varies in nitrogen-based functionalities, exhibiting spectrum properties that encompass varied degrees alkyl substitution (A1-A5), piperazine (A6), oligoamine (A7), guanidinium (A8), and hydroxylation (A9-A10) vary side-chain volume, pattern, hydrophilicity, pK to assess parameter impact on In vitro delivery assays using GFP+ across multiple cell lines reveal amine bulk chemical structure critically affect performance. Using machine learning analysis via SHapley Additive exPlanations (SHAP) 180 formulations (3780 experimental measurements), mapped key relationships between chemistry performance metrics, finding amine-specific binding efficiency was major determinant efficacy, viability, GFP intensity. Micelles stronger capabilities (A1 A7) have higher cellular performance, whereas those intermediate tendencies deliver amount functional per (A2, A10). indicates balancing the strength is crucial hydrophobic bulky pendant groups (A3-A5) tend induce necrosis during delivery, highlighting significance optimization. A7 amphiphile, displaying primary secondary amine, consistently demonstrates highest expression various types vivo achieves high specificity lung tissue upon intravenous administration. Moreover, established strong correlation Multitask Gaussian Process models, underscoring predictive power models anticipating outcomes. Overall, this innovative study integrates advanced data science design, demonstrating pivotal role amine-dependent optimization advancing targeted lungs.

Language: Английский

Machine Learning Reveals Amine Type in Polymer Micelles Determines mRNA Binding, In Vitro, and In Vivo Performance for Lung-Selective Delivery DOI Creative Commons
Sidharth Panda, Ella J. Eaton,

Praveen Muralikrishnan

et al.

JACS Au, Journal Year: 2025, Volume and Issue: unknown

Published: April 14, 2025

Cationic micelles, composed of amphiphilic block copolymers with polycationic coronas, offer a customizable platform for mRNA delivery. Here, we present library 30 cationic micelle nanoparticles (MNPs) formulated from diblock reactive poly(pentafluorophenol acrylate) backbones modified diverse amines. This systematically varies in nitrogen-based functionalities, exhibiting spectrum properties that encompass varied degrees alkyl substitution (A1-A5), piperazine (A6), oligoamine (A7), guanidinium (A8), and hydroxylation (A9-A10) vary side-chain volume, pattern, hydrophilicity, pK to assess parameter impact on In vitro delivery assays using GFP+ across multiple cell lines reveal amine bulk chemical structure critically affect performance. Using machine learning analysis via SHapley Additive exPlanations (SHAP) 180 formulations (3780 experimental measurements), mapped key relationships between chemistry performance metrics, finding amine-specific binding efficiency was major determinant efficacy, viability, GFP intensity. Micelles stronger capabilities (A1 A7) have higher cellular performance, whereas those intermediate tendencies deliver amount functional per (A2, A10). indicates balancing the strength is crucial hydrophobic bulky pendant groups (A3-A5) tend induce necrosis during delivery, highlighting significance optimization. A7 amphiphile, displaying primary secondary amine, consistently demonstrates highest expression various types vivo achieves high specificity lung tissue upon intravenous administration. Moreover, established strong correlation Multitask Gaussian Process models, underscoring predictive power models anticipating outcomes. Overall, this innovative study integrates advanced data science design, demonstrating pivotal role amine-dependent optimization advancing targeted lungs.

Language: Английский

Citations

0