A point-of-care SARS-CoV-2 test based on reverse transcription loop-mediated isothermal amplification without RNA extraction with diagnostic performance same as RT-PCR

Abstract

The global public health crisis and economic losses resulting from the current novel coronavirus disease (COVID-19) pandemic have been dire. The most used real-time reverse transcription polymerase chain reaction (RT-PCR) method needs expensive equipment, technical expertise, and a long turnaround time. Therefore, there is a need for a rapid, accurate, and alternative technique of diagnosis that is deployable at resource-poor settings like point-of-care. This study combines heat deactivation and a novel mechanical lysis method by bead beating for quick and simple sample preparation. Then, using an optimized reverse transcription loop-mediated isothermal amplification (RT-LAMP) assay to target genes encoding the open reading frame 8 (ORF8), spike and nucleocapsid proteins of the novel coronavirus, SARS-CoV-2. The test results can be read simultaneously in fluorometric and colorimetric readouts within 40 min from sample collection. We also calibrated a template transfer tool to simplify sample addition into LAMP reactions when pipetting skills are needed. Most importantly, validation of the direct RT-LAMP system based on multiplexing primers S1:ORF8 in a ratio (1:0.8) using 143 patients' nasopharyngeal swab samples showed a diagnostic performance of 99.30% accuracy, with 98.81% sensitivity and 100% selectivity, compared to commercial RT-PCR kits. Since our workflow does not rely on RNA extraction and purification, the time-to-result is two times faster than other workflows with FDA emergency use authorization. Considering all its strengths: speed, simplicity, accuracy and extraction-free, the system can be useful for optimal point-of-care testing of COVID-19.

Keywords: Bead beating; COVID-19; Point-of-care; RT-LAMP; Rapid testing; SARS-CoV-2.

Graphical abstract

Fig. 1

Discrimination between specific and non-specific RT-LAMP amplification of different primer sets. (A) 2% Gel electrophoresis and colorimetric, (B) fluorometric, and (C) melting curve profiles for specific amplification by primers S1 + ORF8 post-amplification.’1–6’ are SARS-CoV-2 RNA dilutions. ‘N’-negative control, ‘B’-no template control. (D) 2% Gel electrophoresis and colorimetric, (E) fluorometric, and (F) melting curve profiles post-amplification for non-specific amplification using primers Orf1a-HMSe + N2 + S1. Primer Orf1a-HMSe and N2 adapted from other studies [32,33].

Fig. 2

Analytical performance of primers for SARS-CoV-2 and human internal control detection. (A, B) Sensitivity of ORF8, S1, and N-A primers tested in 5-fold serial dilutions of SARS-CoV-2 RNA (1–6). RPP20 primers in dilutions of human gene (1–4), ‘S’- SARS-CoV-2 RNA, ‘N’- negative control, ‘B’- no template control, cp/μL - RNA copies/μL. (C, D) Primer multiplexing in different ratios. (E, F) Specificity of primers against other respiratory pathogens. ‘PC’-positive control, ‘NC’-negative control. (G, H) Probit curve reflects observations from twelve replicates at each dilution. LOD is the least dilution at which 95% of the replicates are detected. Red dotted curves are confidence intervals. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

Effect of sample matrix on the direct RT-LAMP assay based on primer N-A. Tolerance of the RT-LAMP reagents to 1 μL–5 μL of VTM Y (A, B) and VTM X (C, D). Extracted SARS-CoV-2 RNA was used as the template. ‘0’ denotes no VTM used, ‘N’-negative control, ‘B’-no template control. (E, F) 1 μL–5 μL mechanically lysed clinical sample tested to evaluate the tolerance of the system to crude human cell lysates. Here, ‘0’ denotes using purified RNA extracted from the mechanically lysed clinical sample in VTM Y as the template. The dashed line indicates the 30 min (35 min reaction time) threshold time cut-off for this experiment.

Fig. 4

Comparison of mechanical lysis effectiveness for three different beads (Z, Z + C, K) and control (No beads) on selected five positive clinical swab samples. (A) The Tt represents three independent experiments (duplicate technical replicates) and denotes mean and SD. Two-way ANOVA with Tukey multiple comparison was used to compare the differences between groups (Differences between groups with beads Z + C treatment and control are shown). Statistically significant differences are marked ∗, ∗∗, ∗∗∗, indicating p < 0.05, p < 0.01, p < 0.001, respectively. ‘nd’ = not detected. ‘2n’ = no bead treatment for sample 2. (B) Colorimetric and (C) Fluorometric detection for sample 2. Primer N-A was used in this experiment. The corresponding colorimetric and fluorometric results are shown in Appendix A, Fig. S6.

Fig. 5

The impact of mechanical lysis with Zirconium oxide and Chelex on RT-LAMP detection. (A) Effect of vortexing and no vortexing on mechanical lysis of a positive swab sample. (B) Effect of no extraction and subsequent RNA extraction of a mechanically lysed swab sample. (C) RT-PCR detection of extracted RNA from the mechanically lysed sample (D) Mechanical lysis versus other common lysis modalities. (E) RT-PCR detection of extracted RNA from mechanically lysed sample compared with other lysis methods (F) The effect of mechanical lysis time on the release of SARS-CoV-2 RNA. Primer S1:ORF8 (1:0.8) was used in this experiment. Data represent the means ± SD of three technical replicates. The corresponding colorimetric and fluorometric results are shown in Appendix A, Figs. S7 and S8.

Fig. 6

Detection of cell-cultured SARS-CoV-2 spiked in 1x phosphate-buffered saline (PBS) and saliva samples in 10-fold dilution series. (PBS, P1-P7) (Saliva, S1-S7). (A, B, C, D) Detection of extracted RNA from PBS and saliva samples. (E, F, G, H) Detection of crude lysates from mechanically lysed PBS and Saliva samples. (A and E, B and F) Colorimetric and fluorometric readouts, respectively. (C, G) Detection of human internal control gene using RPP20 primers. (D, H) Spearman's correlation of RT-PCR with RT-LAMP for PBS and Saliva samples. Primer S1:ORF8 (1:0.8) was used in this experiment.

Fig. 7

Calibration of template transfer tool. (A) Standard curve for UV–Vis absorbance of GelRed dye against the volumes transferred using a calibrated pipette. The absorbance was measured of 1 μL–5 μL of the dye, each diluted in 500 μL of sterile clean water. The measurements were performed in duplicate at wavelength of 450 nm. Data represented here is mean absorbance against volume. (B) A linear relationship between volumes transferred with the calibrated template transfer tool based on the standard curve against insertion height of transfer tool (Appendix A Table S7). (C) Schematic representation of the calibrated template transfer tool. M, Material; H, Height; D, Diameter. The plastic tool at the insertion height of 2 mm was used for the template transfer.

Fig. 8

Schematic illustration of the direct approach RT-LAMP workflow. After sample collection in VTM, the sample is heat deactivated at 95 °C for 5 min, then mechanically lysed by bead beating method for 5 min. Following lysis, 1 μL of the sample is added to 24 μL of ready-made RT-LAMP master mix using a pipette or the calibrated transfer tool. The reaction is then set up on a portable, battery-operated fluorescent reader (Genie® III; OptiGene Limited, UK). The results can be viewed in a fluorometric and colorimetric output after 30 min of incubation. At the point-of-care, the standard vortex can be replaced by a disposable, battery-powered, miniaturized Omnilyse device (Claremont Biosolutions, USA). All steps should be taken with complete personal protective equipment and all waste disposed off appropriately. The figure was created with BioRender.com and Mindthegraph.com.