Ultra-rapid detection of SARS-CoV-2 in public workspace environments

Abstract

Managing the pandemic caused by SARS-CoV-2 requires new capabilities in testing, including the possibility of identifying, in minutes, infected individuals as they enter spaces where they must congregate in a functioning society, including workspaces, schools, points of entry, and commercial business establishments. Here, the only useful tests (a) require no sample transport, (b) require minimal sample manipulation, (c) can be performed by unlicensed individuals, (d) return results on the spot in much less than one hour, and (e) cost no more than a few dollars. The sensitivity need not be as high as normally required by the FDA for screening asymptomatic carriers (as few as 10 virions per sample), as these viral loads are almost certainly not high enough for an individual to present a risk for forward infection. This allows tests specifically useful for this pandemic to trade-off unneeded sensitivity for necessary speed, simplicity, and frugality. In some studies, it was shown that viral load that creates forward-infection risk may exceed 105 virions per milliliter, easily within the sensitivity of an RNA amplification architecture, but unattainable by antibody-based architectures that simply target viral antigens. Here, we describe such a test based on a displaceable probe loop amplification architecture.

Fig 1. Displaceable probe LAMP to detect SARS-CoV-2.

(A) Classical RT-LAMP utilizes six primers hybridizing to eight regions within the viral genome. These primers form a dumbbell structure through self-hybridization of FIP and BIP and addition of two loop primers improves the amplification rate. To allow simultaneous detection of several targets in real-time, displaceable probe architecture was employed by tagging one of the loop primers with a quencher and supplementing with a partially complementary probe containing a fluorophore tag. (B) Positive results can be analyzed in real-time and process manifests itself as sigmoidal curve as it would be in RT-qPCR using TaqMan probes. (C) End-point fluorescence is then observed through orange filter coupled with blue LED (exc. 470 nm, Firebird Biomolecular Sciences LLC, US).

Fig 2. Limit of detection using DP-RT-LAMP primers using heat-inactivated SARS-COV-2 or human RNA (for internal control).

(A) Real-time analysis of CoV2-W3 primer set (targeting S gene) with LOD of 10 RNA copies/assay. (B) End-point visualization of LAMP products with primer set CoV2-W3. (C) Real-time analysis of CoV2-v2-4 primer set (targeting N gene) with LOD of 25 RNA copies/assay. (D) Real-time analysis of internal control RNaseP-2 primer set (targeting human RNase P gene) with LOD of 44 copies of human RNA/assay. (E) Time to threshold (Tt) values of each LAMP primer set was determined for each target copy number/assay and similar values were obtained when compared to RT-qPCR test (Ct values were also converted to their corresponding Tt values for convenience).

Fig 3. Sampling work-flow and results output.

(A) Dry nasal swabs were used as sampling method. Swabs were first eluted in a sample preparation buffer and aliquot from that was added into RT-LAMP mixture. End-point results were visualized using blue LED and orange filter. (B) Direct saliva was mixed with a sample preparation buffer briefly and aliquot from that was added into RT-LAMP mixture. End-point results were visualized using the same method for nasal swab sampling. (C) Q-paper was combined with saliva and Q-paper coated with saliva was directly introduced into RT-LAMP mixture without further manipulation. End-point fluorescent signal was visualized using blue LED and orange filter. Note that the square of Q-paper is observable, but does not compromise the real-time or end-point analysis. (D) In addition to end-point visualization, RT-LAMP experiments were also run in real-time using Genie^® II (Optigene, UK) which can operate on battery therefore enabling its use in low-resource settings.

Fig 4. Further evaluation of presently preferred sampling methods and sensitivity analysis with contrived samples using heat-inactivated SARS-CoV-2 template from BEI.

(A) Varying amounts of RNA was spiked into nasal swab samples from healthy individuals. 200 copies of RNA were detected with consistency and RNase P gene was used as sampling control. (B) and (C) Varying amounts of RNA was spiked into saliva samples or saliva that was deposited onto Q-paper, respectively. 200 copies of RNA were detected with consistency and RNase P gene was detected successfully. (D) Mean Tt values in minutes and numbers of positive results versus total number of samples were displayed in the table.

Fig 5. Multiplexed detection of SARS-COV-2 RNA and RNase P (internal control).

(A) Varying amounts of heat-inactivated SARS-CoV-2 (BEI resources) spiked with human RNA (440 cp). Fluorescence signals from three channels were recorded every 30 seconds using LightCycler^® 480. Channel 483–533 is specific for SARS-CoV-2 RNA, channel 523–568 can detect signals from both targets (ladder formation manifests itself), and channel 558–610 is specific for RNase P. Corresponding Tt values were shown on the table. (B) 10⁴, 10³ and 10² copies of SARS-CoV-2 RNA was spiked into processed nasal swab and saliva samples and analyzed simultaneously.