Ethics Statement: UHL NHS patient ON swab samples used in this study were collected in the context of routine clinical patient care. RT-LAMP analyses performed on residual de-identified patient material served as post-RT-PCR assay validation. Fast track ethical approval was obtained saliva tests allowing us to assess the performance of saliva direct RT-LAMP (no RNA extraction) against RT-LAMP on extracted RNA from matched saliva sample. (REC reference: 20/EE/0212).
University of Leicester Asymptomatic Screening Programme
Participants wishing to access the COVID-19 screening programme required registration at the primary care Leicester Victoria Park Health Centre in order to generate a pathology request form. Participants self-swabbed (throat and lower nasal cavity) at a supervised screening venue using Miraclean swabs placed into PrimeStore Molecular Transport Medium for viral inactivation and RNA stabilisation at room temperature (Longhorn Vaccines and Diagnostics). Within a 24-hour period, total nucleic acid extraction was followed by RT-LAMP (65°C for 20 minutes) targeting the Orf1a (Orf1a-HMSe primer design) plus an internal ACTB control (NEB primer design). Data was uploaded to the pathology iLab system and NHS laboratory RT-PCR testing confirmed any positive RT-LAMP results, feeding into the national track and trace system. Results were reported back to participants (< 48 hours post sample collection) via SMS (for negative result) or phone call (for positive results) from the Victoria Park Health Centre. Figure 1 details the programme workflow.
Control RNA: Synthetic SARS-CoV-2 RNA at a concentration of 1×106 RNA copies per microliter was purchased from Twist Bioscience and diluted appropriately in nuclease free water (Twist Synthetic SARS-CoV-2 RNA Control 1 [MT007544.1] - SKU: 102091 and Control 2 [MN908947.3] - SKU: 102024). Negative control RNA from related Betacoronavirus 1 (Strain OC43) and non-related Influenza A (H1N1) was purchased from ATCC. Total human RNA purchased from Invitrogen (4307281).
Swab sample collection and RNA extraction
Standard ON swabs from hospital inpatients were collected using PHE-approved flocked swabs placed into viral transport media (Virocult / VTM-M4RT). RNA extraction (on 200 µl of inactivated sample mixed with 265 µl Binding Solution) was carried out using the MagMAX Viral/Pathogen Nucleic Acid Isolation kit (MVPII, Thermofisher) on the KingFisher Flex Purification System. Samples were stored at 4°C short term prior to RT-LAMP and diagnostic RT-PCR confirmation at the University Hospital Leicester NHS Trust (UHL).
Saliva Collection and RNA extraction
Research nurse assisted UHL inpatient saliva sample collection was performed at least 30 minutes after a meal, taking of oral medication or brushing of teeth. Pre-collection, patients rinsed their mouth with water and then waited 10 minutes before up to 5 ml of saliva was collected in universal collection vessels and stored at 4°C for up to three days prior to RNA extraction (as detailed above). Purified RNA was stored at -20°C prior to RT-LAMP (performed using the Rotor-Gene Q platform) or RT-PCR processed through the diagnostic service in the University Hospitals of Leicester NHS Trust.
NHS real-time RT-PCR
Purified nucleic acid from ON swab samples or saliva was reverse transcribed into cDNA and amplified using a CE marked, locally validated commercially available kit targeting the E and S-gene sequence regions (RealStar® SARS-CoV-2 RT-PCR kit, Altona Diagnostics GmbH, Hamburg, Germany). Residual RNA from samples processed through the diagnostic RT-PCR service in the UHL NHS Trust with corresponding Ct value were frozen at -80°C until required for validation of RT-LAMP.
RT-LAMP primers
Primers targeting the Orf1a (Orf1a-HMSe) were designed by Rabe and Cepko 9. Primers targeting the nucleocapsid (N), envelope gene (E) and internal human beta-actin internal control (ACTB) were designed by NEB 10. Primer sequences are listed in supplementary Table 3. HPLC grade purification primers were synthesised by Merck and reconstituted in nuclease free water. Individual RT-LAMP primer sets were prepared as 20 times final concentration stock, with final assay concentrations of 0.2 µM F3/B3, 1.6 µM FIP/BIP and 0.4 µM LoopF/Loop B.
Fluorescent RT-LAMP
Reactions contained 1 X WarmStart® LAMP Master Mix (E1700) supplemented with 1 X fluorescent dye (NEB dye provided with E1700 master mix), 0.02 U/µL Antarctic Thermolabile UDG (NEB), 700 µM Thermolabile dUTP (NEB) and 1 X standard concentration LAMP primers. Unless otherwise stated, reactions were prepared to a final reaction volume of 25 µl using nuclease free water and incubated at 65°C using a either a StepOnePlus (Applied Biosystems) or Rotor-Gene Q (Qiagen) thermoclycler for up to 30 cycles (unless otherwise stated), with signal acquisition every 60 s.
Colorimetric RT-LAMP
Reactions contained 1 X WarmStart® Colorimetric LAMP Master Mix (M1800; NEB) supplemented with 1 x EvaGreen (Biotium), 0.02 U/µL Antarctic Thermolabile UDG (NEB), 700 µM Thermolabile dUTP (NEB), 1 X standard concentration LAMP primers and 40 mM guanidine chloride solution (Sigma G3272, pH adjusted to pH ~ 8). Reactions were prepared to final volume of 25 µl using nuclease free water, incubated at 65°C for 30 minutes on a StepOnePlus thermocycler (Applied Biosystems). The colour of finished reactions was recorded using an office flatbed scanner.
Saliva direct RT-LAMP
On the day of sample collection, 50 µl of saliva was mixed with 50 µl of MicroLYSIS buffer (Clent Scientific), heated at 95°C for 10 minutes to achieve viral inactivation, placed on ice and processed via fluorescent RT-LAMP in under 20 minutes. All saliva assays were performed using the Rotor-Gene Q platform.
Statistical analysis
Time to positive (TTP in minutes) served as a surrogate for RT-PCR Ct and a semi-quantitative measure of viral RNA concentration. Additional product specificity checks were provided by melt curve analysis within acceptance range (2 degrees either side of the mean Tm determined during assay validation using patient samples). Regression calculations determining reaction efficiency incorporated data from concentrations where three or more values reported a TTP in under 25 minutes. Reactions were considered as negative with a TTP above 25 min. Reaction efficiency is calculated by E=-1 + 10(− 1/slope). Validation data using synthetic material represent the average of two independent experiments, performed in quadruplicate, presented as mean TTP ± S.E.M. Validation data using residual patient RNA is a single reaction performed in parallel reactions targeting the Orf1a, N + E and ACTB total RNA control sequence. MedCalc® Scientific Software was used for diagnostic test evaluation to determine the diagnostic sensitivity (DSe) referring to the proportion of known positive samples that tested positive in the assay and diagnostic specificity (DSp) referring to the proportion of samples from known negative reference samples that test negative in the assay.