Fast detection of SARS-CoV-2 RNA via the integration of plasmonic thermocycling and fluorescence detection in a portable device

Abstract

The diagnosis of severe acute respiratory syndrome 2 (SARS-CoV-2) infection by quantitative PCR with reverse transcription (RT-qPCR) typically involves bulky instrumentation in centralized laboratories and an assay time of 1-2 h. Here, we show that SARS-CoV-2 RNA can be detected in 17 min via a portable device integrating reverse transcription, fast thermocycling (via plasmonic heating through magneto-plasmonic nanoparticles) and in situ fluorescence detection following magnetic clearance of the nanoparticles. The device correctly classified all nasopharyngeal, oropharyngeal and sputum samples from 75 patients with COVID-19 and 75 healthy controls, with good concordance in fluorescence intensity with standard RT-qPCR (Pearson coefficients > 0.7 for the N1, N2 and RPP30 genes). Fast, portable and automated nucleic acid detection should facilitate testing at the point of care.

Figure 1.. nanoPCR assay schematics for COVID-19 diagnostics.

(a) Target RNA regions for nanoPCR tests. N1, N2 genes are for SARS-CoV-2 detection and RPP30 gene serves as a control for human sample confirmation. (b) Accelerated nanoPCR diagnostic flow for SARS-CoV-2 detection: i) 3 min of RNA extraction using a disposable RNA-prep kit; ii) 12 min of reverse transcription (RT) and polymerase chain reaction (PCR) by magneto-plasmonic thermocycling; iii) 3 min of detection and diagnosis via magnetic fluorescence switch. (c) Synchronized Ferris wheel for multi-sample loading thermocycling. Syncing laser illumination with Ferris wheel rotation enables the processing of multiple samples in turn. (d) A compact nanoPCR instrument for point-of-care (POC) operation.

Figure 2.. Magneto-plasmonic nanoparticles (MPNs).

(a) TEM images of MPNs with a core-shell (Zn_0.4Fe_2.6O₄-Au) structure. The core size was 16 nm and the shell thickness 12 nm (left). Elemental mapping of Au and Fe with EDS showed area specific distribution of core and gold shell structure (right). (b) UV-Vis absorption cross-section spectrum of MPN measured with integrating sphere. At the shell thickness of 12 nm, the highest absorbance occurred at 535 nm (triangle – not shown here). (c) Simulation of electric field (via finite-difference time-domain (FDTD) method) at 535 nm illumination. Intense electric field was confined at Au surfaces. (d) Linear profile of maximum electric field enhancement factor (E/E₀) (dotted line in (c)). (e) Magnetization curves of single and clustered form of MPNs measured by vibrating sample magnetometer (VSM). Clustered form of MPN was realized by dipolar interaction under magnetic field application (100 T/m). MPNs and clusters both exhibited superparmagnetism, but clustered MPN shows ~1,000 times higher magnetic moment. (f) Hydrodynamic sizes of MPNs before and after clusterization measured by dynamic light scattering. Size increased from 48 nm to 500 nm. Clustered MPNs experience strong pulling force under external magnetic fields to be rapidly cleared from the solution.

Figure 3.. MPN-induced plasmonic heating.

(a) Thermal images of plasmonically heated MPNs solution. Solution temperature changed from 25°C to 90°C within a few seconds upon laser illumination. Light source of 1000 mW at λ= 532 nm; solution volume, 10 μl; MPN concentration, 2.6 × 10¹¹ particles/mL. (b) Temperature profile of MPN solutions with different wavelengths of light illumination. 532 nm laser diode matched to the maximum absorption of MPN, and thereby faster heating was induced. (c) Effect of MPN concentration on temperature profile (dotted line: theoretical estimation, circle: measured). The maximum temperature rise observed at 2.6 × 10¹¹ particles/mL. The higher concentration suffered and heating effect becomes diminished. (d) High-speed plasmonic thermocycle profile of MPN solution (7 cycles/min, 58 to 95 °C). (e) Single thermocycle of MPN solution (red dotted square in (d)). Rising rate 13.17 °C/sec; cooling rate 4.94 °C/sec. The temperature deviations (right) were from xxx cycles at 58 and 95 °C. The values were < 1.64% (compared to what?). (f) Schematic of synchronized Ferris wheel system for multi-sample thermocycling. Sample rotation and laser illumination were synced to heat three samples in turn. (g) Temperature profiles of three samples set in the wheel. Individual heating profiles are interleaved such that total cycling time was not affected.

Figure 4.. nanoPCR detection of SARS-CoV-2 genes.

(a) Temperature profile of nanoPCR used in COVID-19 RT-PCR diagnosis. 42°C incubation time for RT (5.5 min) and PCR amplification (40 cycles, < 6.5 min), followed by signal detection steps (3 min). The Whole RT-PCR process was complete within 15 min. (b) Gel electrophoresis image of PCR products (N1, N2, RPP30 genes with or without primer templates) from 15 min nanoPCR and 2 hr conventional benchtop thermocycler. All bands are detected identical for both methods. (c) Magnetic fluorescence switch (MFS) for in-situ detection of amplicons. Photographs and fluorescence images (FAM dye) of an assay mixture before and after 3 min MFS application. (d) N1 gene fluorescence signal changes during MFS application. 50% signal recovery was achieved at 3 min. (e) Detection of N1, N2, and RPP30 genes as well as non-target control (NTC) gene before/after MFS application. (f) Evaluating limit of detection (LOD) of nanoPCR with different amounts of target genes. Detection threshold was set at 3 × standard deviation of signal from a blank. LOD was 3.2 copies of gene/μL. (g) Specificity evaluation for different strains of corona virus (i.e. N1 and N2 genes from SARS-CoV-2, SARS-CoV, and MERS-CoV).

Figure 5. Point-of-care (POC) testing of clinical samples with nanoPCR for COVID-19 diagnosis

(a) Overall schematics of clinical diagnostics of SARS-CoV-2 in POC setting. After swab sample collection, RNA preparation is done with RNA-prep kit (3 min), followed by high-speed nanoPCR amplification and fluorescence detection (15 min). 18-minutes completion from sampling to COVID-19 diagnosis. (b) nanoPCR testing of clinical samples. Twenty-five COVID-19 positive samples and another 25 COVID-19 negative controls were used. COVID-19 status of samples was independently confirmed at a hospital clinical laboratory. nanoPCR reported high N1 and N2 signal in all COVID-19 positive samples; the corresponding signal was absent in control samples. All samples were positive in RPP30. (c) Evaluating analytical concordance between nanoPCR and quantitative RT-PCR (qRT-PCR). The results were linearly correlated (R² = 0.8665). Each LOQ of both qRT-CPR and nanoPCR is plot on each axis (blue broken line)