Mechanisms of Adeno-Associated Virus Serotype 9 Vector Characterization and Quality Control through Solid-State Nanopores

Adeno-associated virus (AAV) vectors are excellent gene-delivery carriers in gene therapy; however, improperly packaged capsids produced during manufacturing can reduce potency and raise safety concerns. We introduce a machine-learning-assisted, low-cost, label-free nanopore sensing platform with single-particle resolution to enhance AAV quality control. Using solid-state nanopore (SSN) devices on Si_xN_y membranes, we optimized in vitro conditions for AAV9 detection and classification. We observed pH-dependent capsid denaturation under strong alkaline conditions. Buffer-specific, selective translocation of emptyAAV9 capsids from cargo-loaded samples enabled clear discrimination and revealed potential avenues for in situ filtration. We also observed distinct translocation behaviors between vectors encapsulating single-stranded DNA and those encapsulating self-complementary DNA. In addition, unsupervised clustering algorithms demonstrated high accuracy in distinguishing capsids with truncated genomes from those with full genomes, further facilitating AAV production quality. These findings support practical avenues for AAV filtration and analysis, providing a basis for label-free, high-throughput, precise, and scalable quality control in AAV vector manufacturing.