Picomolar Fingerprinting of Nucleic Acid Nanoparticles Using Solid-State Nanopores

Abstract

Nucleic acid nanoparticles (NANPs) are an emerging class of programmable structures with tunable shape and function. Their promise as tools for fundamental biophysics studies, molecular sensing, and therapeutic applications necessitates methods for their detection and characterization at the single-particle level. In this work, we study electrophoretic transport of individual ring-shaped and cube-shaped NANPs through solid-state nanopores. In the optimal nanopore size range, the particles must deform to pass through, which considerably increases their residence time within the pore. Such anomalously long residence times permit detection of picomolar amounts of NANPs when nanopore measurements are carried out at a high transmembrane bias. In the case of a NANP mixture, the type of individual particle passing through nanopores can be efficiently determined from analysis of a single electrical pulse. Molecular dynamics simulations provide insight into the mechanical barrier to transport of the NANPs and corroborate the difference in the signal amplitudes observed for the two types of particles. Our study serves as a basis for label-free analysis of soft programmable-shape nanoparticles.

Keywords: DNA cube; RNA and DNA nanotechnology; RNA ring; nanopore sensing; nucleic acid nanoparticles.

Figure 1

Assembly and visualization of nucleic acid nanoparticles (NANPs). Cartoon and 3D models (each strand colored differently) that define the structure of DNA cubes (a) and RNA rings (b). Below each structure, AFM images of the assembled structures are shown. (c) Ethidium bromide total staining native-PAGE of rings, cubes, and mixture of rings and cubes. (d) Schematics of a nanopore drawn approximately to scale with the two NANPs. Inset: Transmission electron micrograph of a 9 nm pore fabricated using electron beam irradiation of a 50 nm thick silicon nitride membrane (scale bar: 5 nm).

Figure 2

Deformation-controlled transport of NANPs. (a) Normalized current traces obtained for RNA rings interacting with a 9 nm nanopore at 200, 400, 600, 800, and 1000 mV (for 400 and 600 mV data, open-pore segments were excised to allow more events to be seen). (b) Mean dwell time versus applied bias for RNA rings (dashed line is a trendline fit to guide the eye). Vertical lines for each data point represent ±1 standard deviation in the corresponding dwell time distributions. (c) Mean fractional blockade versus applied bias (dashed line is a linear fit). Inset cartoons are perceived interpretations of the increasing fractional blockades with increasing voltage. (d) Recapture of an RNA ring. Switching polarity of the voltage bias immediately after a long ionic current blockade produced a similar amplitude opposite polarity blockade event, confirming nanopore translocation of NANPs and further suggesting that NANPs retain their structure after translocation (inset scale bar: 300 µs). Experiments performed with 400 mM KCl (10 mM Tris, 2 mM MgCl₂, pH 7.9).

Figure 3

Concentration limit of NANP detection. (a–h) Current traces recorded from samples containing RNA rings at different concentrations. All measurements were carried out using a 9 nm pore at 1 V applied bias; the data were low-pass filtered at 500 kHz. Shaded regions in panels a–c highlight occurrence of shallow events. (i) Capture rate versus concentration of RNA rings. Red markers: capture rate based on total number of events. Black markers: capture rate after omission of the shallow events with fractional blockade less than 15%. Experiments performed with 400 mM KCl (10 mM Tris, 2 mM MgCl₂, pH 7.9).

Figure 4

Discrimination of RNA rings from DNA cubes. (a) Current traces measured from a binary mixture of DNA cubes and RNA rings (1 nM ring; 1.4 nM cube). The translocation experiments were performed using a 9.5 nm diameter pore, with a 500 mV transmembrane voltage; data were low-pass filtered at 500 kHz. (b) Close view of the current trace indicates two distinct blockade levels associated with each nucleic acid nanoparticle. (c) Scatter plot of the fractional blockade versus dwell time measured at 500 mV applied bias displays two distinct populations corresponding to the DNA cubes (n = 422) and the RNA rings (n = 376). (d) Scatter plot similar to (c) obtained at an applied bias of 800 mV. In this panel, data were processed using a 1 MHz low-pass filter (n_cube = 573, n_ring = 498). Experiments performed with 400 mM KCl (10 mM Tris, 2 mM MgCl₂, pH 7.9).

Figure 5

MD simulations of NANPs’ translocation through a solid-state nanopore. (a) Typical simulation system containing a nanopore (gray), a DNA cube (colors), water (semitransparent surface), and ions (spheres). The diameter of the nanopore constriction is 9 nm. (b) Snapshots illustrating a simulated translocation of a DNA cube under a 500 mV transmembrane bias. Water and ions are not shown for clarity. (c,d) Representative conformations of a DNA cube (c) and an RNA ring (d) trapped at the nanopore constriction under a 200 mV transmembrane bias. (e) Percentage of the open-pore current reduced by the nanoparticles trapped in the nanopore. Several independent simulations were performed for each particle at each bias condition differing by the initial orientation of the particles with respect to the nanopore (see SI Figure S6). The blockade currents were determined after the particles reached a stable conformation within the nanopore (see SI Figure S7 and Table S1). The error bars represent the propagated standard error of 800 ps block averaging. A horizontal dashed line denotes the average blockade currents produced by DNA cubes and RNA rings.