The nanopore force spectroscopy provides the most direct way to understand mechanical
properties of soft/rigid nanoparticles.
Soft, membrane-encased nanoparticles, such as endosomes, liposomes, and viral particles, have significant impact on a variety of physiological processes ranging from endo- and exocytosis, drug delivery, and infection. Profiling their mechanical and payload properties are critical to understanding their role and efficacy in biological systems. Mechanical properties, such as bending rigidity and surface tension, are crucial to gauge their infectivity stage, transport of foreign material to host cells, and receptor-reporter interactions, whereas payload properties are essential for dose-dependent assays and understanding their medicinal/biological/pathological function(s). The current instruments/techniques used for these properties have one or more drawbacks: low-throughput, high overhead cost, highly trained personnel, tailored facilities, and tedious sample preparations. Additionally, most of the methods would either profile mechanical or payload properties, but not both. To address these shortcomings, our ongoing research aims to develop an easy-to-use, bimodal nanopore-based bioanalytical tool for comprehensive profiling of biomimetic lipid nanoparticles (with their payload) of interest. Both optical (single-point confocal spectroscopy) and electrical (resistive pulse sensing) signals can be obtained simultaneously for mechanical characterization and cargo determination. Furthermore, in collaboration with Prof. Prashanta Dutta at Washington State University, we are developing a probabilistic (Bayesian) machine learning framework to identify the cargo loading from experimental data by modifying an existing deep neural network. The transfer learning will be used to train the deepest layers of the network from the limited data obtained from the nanopore experimental study. Our bimodal nanopore platform is able to operate at sub-picomolar concentration limits and requires very small sample volumes (<50uL dilute sample) – a substantial benefit when vector preparations are expensive and labor intensive.
Recent advancements in targeted gene delivery and personalized medicine are enabling disease treatments that have hitherto been unattainable. For example, gene therapy driven by viruses like the adeno-associated virus (AAV) has already received FDA approval (Luxturna) to treat hereditary blindness. Critically, the effectiveness of such treatments demands fast and sensitive sensing methods to discriminate cargo-filled capsids from empty ones. Under- or over-estimation of the empty capsids could lead to wrong dosing of the therapeutic that could lead to ineffective or even catastrophic outcomes. Blind studies of vector titers using qPCR and ELISA have raised concerns about their ability to quantify the fraction of empty capsids with any accuracy. To address this important need in the field of gene therapy, we are developing a new solid-state plasmonic nanopore technology that can distinguish between empty and DNA-filled AAV capsids based on their bimodal optical-electrical deformation characteristics that are measured repeatedly by an automated recapture protocol. In recent work, we show the ability of electrical nanopore sensing to discriminate between nanoparticles of different rigidity levels through translocation and recapture measurements. These days we are taking on the challenge of discriminating between different levels of rigidity in empty versus DNA-loaded AAVs through increased sensitivity and specificity offered by bimodal optical-electrical nanopore sensing coupled to a recapture protocol. Our technology could increase significantly the quality of therapeutic carrier preparations by helping sort AAVs by acting as a nano-sieve based on their deformability and charge density, which are affected by their DNA cargo content.