Addressable nanoantennas with cleared hotspots for single-molecule detection on a portable smartphone microscope

The study addresses the challenge of detecting single-molecule fluorescence without expensive, high-NA optical instruments. Conventional single-molecule detection requires costly equipment due to extremely weak nanoscale signals. While plasmonic nanoantennas can enhance fluorescence and DNA origami enables precise emitter placement, previous designs had limited hotspot accessibility and modest enhancement (e.g., ~7.3-fold with monomer antennas), insufficient for detection using low-NA, smartphone-based microscopes that typically require signals equivalent to ≥16 emitters. The goal is to create DNA-origami–scaffolded nanoantennas with cleared, addressable hotspots (NACHOS) that both preserve high fluorescence enhancement and allow integration of bioassays, enabling sensitive and specific single-molecule detection (e.g., of antibiotic-resistance gene fragments) on portable, low-cost smartphone microscopes.

The paper situates its work within two approaches to high-sensitivity detection: molecular amplification (e.g., isothermal amplification) and physical signal enhancement (plasmonic fluorescence enhancement). Prior work demonstrated plasmonic nanoantennas and DNA origami for emitter positioning but suffered from stringent gap requirements and obstructed hotspots, limiting bioassay integration. A nanoantenna-based Zika DNA assay with a quenched hairpin in a monomer configuration yielded an average enhancement of ~7.3-fold, inadequate for low-NA optics. Smartphone-based fluorescence microscopy benchmarks indicated a need for signals equivalent to at least ~16 single emitters for detection. The literature highlights the potential of plasmonic enhancement, the utility of DNA origami for precise nanogap engineering, and the unmet need for designs that combine strong enhancement with assay accessibility and affordability.

Design and assembly of NACHOS: A three-dimensional DNA origami structure was designed (caDNAno) using an M13mp18-derived scaffold and staple strands, forming a cross-like rigid base (~35×33 nm) with six biotin-modified staples for NeutrAvidin-mediated immobilization on BSA-biotin coated coverslips. Two ~83 nm pillars each carry six A20 strands to bind 100 nm silver nanoparticles functionalized with T20-SH DNA in a zipper-like geometry, creating a plasmonic hotspot (~12 nm gap) between the nanoparticles and pillar bifurcation. TEM confirmed assembly. Fluorescence enhancement characterization: An Alexa Fluor 647-labeled staple was positioned in the hotspot. Single-molecule confocal measurements compared NACHOS (two 100 nm Ag NPs) to reference origami without nanoparticles, analyzing intensity, single-step bleaching, and SBR. Sandwich DNA detection assay: For diagnostics, three capture strands protruding into the hotspot were integrated to detect a 34-nt DNA segment specific to the OXA-48 carbapenemase gene (Klebsiella pneumoniae). Upon hybridization with the target, a 17-nt Alexa Fluor 647-labeled imager strand binds, placing the dye in the hotspot. The origami base carried a single green reference dye (ATTO 542) for colocalization. Incubation was typically 2 h at 37 °C with 2 nM target and 6 nM imager; binding was also observed after 15 min. Controls included imager-only incubation and target strands with 1–3 nt mismatches. Quantification: Confocal scans measured colocalized green/red spots (binding yield), enhancement versus reference, and single-molecule photobleaching step distributions to count imagers per hotspot. Serum compatibility: The assay was repeated in heat-inactivated human blood serum spiked with target and imager, maintaining identical conditions. Smartphone microscope construction and measurements: A portable, battery-driven system used a 638 nm, 180 mW laser focused at ~45° onto the sample (illumination area ~150×200 µm²), a low-cost objective lens (NA 0.25; ~$8; ~1.2 µm resolution in red) for collection, bandpass filtering, and a Huawei P20 monochrome smartphone camera for detection and processing. Samples were sealed; ROXS buffer with enzymatic oxygen removal was used to enhance Alexa 647 photostability. Movies (80 ms/frame typical) were recorded and analyzed (ImageJ macro, Origin). Controls included samples without nanoparticles and slides incubated with nanoparticles only to assess scattering. Additional methods: Detailed protocols for DNA origami folding/purification, silver nanoparticle DNA functionalization, magnetic bead-based solution-phase assembly for TEM, coverslip functionalization, confocal setup parameters (laser powers, detection, analysis), and smartphone data analysis are provided.

- Cleared-hotspot NACHOS preserve strong enhancement: Alexa 647 incorporated directly yielded up to 417-fold enhancement (average 74 ± 3-fold). SBR improved to 361 ± 35 versus 7.4 ± 0.9 in reference. - Sandwich DNA assay (OXA-48 target): Enhancement up to 461-fold (average 89 ± 7-fold) with single-step bleaching confirming single emitters in hotspot. Non-amplified background from nonspecific binding remains low. - Binding yields: Fraction of origami with target/imager bound was 66% in NACHOS and 84% in reference (higher nonspecific imager sticking in reference likely contributes to higher apparent yield). False positives (imager-only control) ~2.5%. Mismatches (1–3 nt) reduced colocalization, indicating sequence selectivity. - Photobleaching step analysis (number of imagers per hotspot): ~60% single, ~30–33% double, and ~8–11% triple bleaching steps for both NACHOS and reference, indicating unobstructed hotspot accessibility. - Serum robustness: In heat-inactivated human blood serum, performance matched buffer: enhancement up to 457-fold (average 70 ± 4), similar binding yields and bleaching step distributions, demonstrating stability and assay compatibility in complex fluids. - Smartphone microscope single-molecule detection: First demonstration of single-molecule fluorescence and DNA sandwich assay on a portable smartphone microscope with non-specialized, low-NA optics. Observed single-step and multi-step photobleaching, blinking, and disappearance of spots over time. Measured SBR 25 ± 2 and SNR 3.8 ± 0.2. Controls without nanoparticles showed no signal; occasional dim non-bleaching spots from silver aggregate scattering were noted in nanoparticle-only controls. - Practicality: Low-cost collection optics (~$8 objective; system components ~€4200 prototype, potentially <€1000 in scale), per-sample preparation cost <€5. NACHOS exhibited stability over extended periods (single-molecule data maintained up to 13 weeks).

The study demonstrates that redesigning DNA-origami nanoantennas to clear and address the hotspot enables both high fluorescence enhancement and assay integration, overcoming the historic trade-off between hotspot accessibility and enhancement. The strong amplification restricted to the hotspot allows specific single-molecule signals to stand out against unavoidable background and nonspecific adsorption, which is not amplified. Binding yields and bleaching-step distributions confirm that attaching two 100 nm silver nanoparticles does not obstruct target access. Comparable performance in human serum underscores robustness and applicability in realistic sample matrices. Crucially, the amplified signals enable, for the first time, single-molecule detection and readout of a diagnostic DNA assay on a portable, battery-powered smartphone microscope with low-NA optics, advancing the feasibility of affordable point-of-care single-molecule diagnostics.

NACHOS—DNA-origami nanoantennas with cleared, addressable hotspots—achieve high fluorescence enhancement without compromising hotspot accessibility, enabling single-molecule detection of disease-specific DNA both on benchtop confocal setups and on a portable smartphone microscope. The platform is robust in complex biological fluids, low-cost, and compatible with straightforward sample preparation. These results provide a pathway toward democratizing single-molecule detection for point-of-care diagnostics. Future work could focus on optimizing capture strand design for enhanced specificity and mismatch discrimination, multiplexed target detection within single nanoantennas, integration into disposable microfluidic cartridges, broader panels of biomarkers (RNA, proteins via aptamers), improved photostability and dye choices, standardized manufacturing to reduce cost and variability, and automated on-device analysis software for clinical workflows.

- Enhancement heterogeneity: Broad enhancement distributions reflect variability in nanoparticle size, shape, and orientation, and presence of monomer antennas. - Smartphone performance constraints: Low NA and modest SNR (3.8 ± 0.2) limit resolution and may challenge highly crowded samples; careful dilution is required. - Photophysics: Alexa 647 required ROXS and oxygen scavenging for clear single-step bleaching; dye photostability may limit some use cases. - Specificity and nonspecific binding: Although false positives were low (~2.5%) and mismatches reduced binding, capture/imager designs could be further optimized to improve selectivity; reference structures showed higher nonspecific sticking. - Controls indicate occasional scattering from silver aggregates in nanoparticle-only samples. - Prototype cost and assembly: While optics are inexpensive, the current prototype system cost (~€4200) and manual fabrication may limit immediate deployment; scaling is anticipated to reduce costs. - Generalizability to diverse targets and real clinical specimens beyond spiked serum remains to be validated, including sample preparation and matrix effects.