key: cord-0934727-c6dg4zsa
authors: Howson, E.; Kidd, S.; Sawyer, J.; Cassar, C.; Cross, D.; Lewis, T.; Hockey, J.; Rivers, S.; Cawthraw, S.; Banyard, A.; Anderson, P.; Rahou, S.; Andreou, M.; Morant, N.; Clarke, D.; Walsh, C.; Laxman, S.; Houghton, R.; Slater-Jefferies, J.; Costello, P.; Brown, I.; Cortes, N.; Godfrey, K.; FOWLER, V.
title: Preliminary optimisation of a simplified sample preparation method to permit direct detection of SARS-CoV-2 within saliva samples using reverse-transcription loop-mediated isothermal amplification (RT-LAMP)
date: 2020-07-21
journal: nan
DOI: 10.1101/2020.07.16.20155168
sha: 974834a78b23cbf923b1f6069a264ab23b451cd1
doc_id: 934727
cord_uid: c6dg4zsa

We describe the optimization of a simplified sample preparation method which permits rapid and direct detection of SARS-CoV-2 RNA within saliva using reverse-transcription loop-mediated isothermal amplification (RT-LAMP). Treatment of saliva samples prior to RT-LAMP by dilution 1:1 in MucolyseTM, followed by dilution (within the range of 1 in 5 to 1 in 40) in 10% (w/v) Chelex 100 Resin and a 98oC heat step for 2 minutes enabled detection of SARS-CoV-2 RNA in all positive saliva samples tested, with no amplification detected in pooled negative saliva. The time to positivity for which SARS-CoV-2 RNA was detected in these positive saliva samples was proportional to the real-time reverse-transcriptase PCR cycle threshold (CT), with SARS-CoV-2 RNA detected in as little as 05:43 (CT 21.08), 07:59 (CT 24.47) and 08:35 (CT 25.27) minutes, respectively. The highest CT where direct RT-LAMP detected SARS-CoV-2 RNA was 31.39 corresponding to a 1 in 40 dilution of a positive saliva sample (1:1 in MucolyseTM) with a starting CT of 25.27. When RT-LAMP was performed on pools of SARS-CoV-2 negative saliva samples spiked with whole inactivated SARS-CoV-2 virus, RNA was detected at dilutions spanning 1 in 5 to 1 in 160 representing CTs spanning 22.49-26.43. Here we describe a simple but critical rapid sample preparation method which can be used up front of RT-LAMP to permit direct detection of SARS-CoV-2 within saliva samples. Saliva is a sample which can be collected non-invasively without the use of highly skilled staff and critically can be obtained from both healthcare and home settings. Critically, this approach overcomes both the requirement and validation of different swabs and the global bottleneck in obtaining RNA extraction robots and reagents to enable molecular testing by PCR. Such testing opens the possibility of public health approaches for effective intervention to control the COVID-19 pandemic through regular SARS-CoV-2 testing at a population scale, combined with isolation and contact tracing for positive cases.

The COVID-19 pandemic caused by the SARS-CoV-2 virus poses a profound global threat to 61

Optimisation of Direct RT-LAMP for detection of SARS-CoV-2 RNA in saliva was performed using three 114 SARS-CoV-2 positive saliva samples collected from symptomatic patients at Hampshire Hospitals NHS 115 Foundation Trust (HHFT) (n=1) and University Hospital Southampton (UHS) (n=2) who had previously 116 had rRT-PCR positive SARS-CoV-2 positive naso-pharyngeal samples. An additional 15 SARS-CoV-2 117 negative saliva samples collected from asymptomatic UHS healthcare staff were used to prepare a 118 pooled sample for specificity analysis. For spiking experiments, one pool of 25 SARS-CoV-2 negative 119 saliva samples from asymptomatic UHS staff, and a second pool of 5 SARS-CoV-2 negative saliva 120 samples also from asymptomatic UHS staff, were used to prepare pooled samples for spiking with 121 whole inactivated virus (SARS-CoV-2 at ~1x10 5 TCID50/ml was inactivated using beta-propriolactone 122 [BPL] ). Collection of saliva involved the patient providing a fresh saliva sample into a 10 ml universal 123 container. Each positive saliva sample was diluted 1:1 in Mucolyse TM (active ingredient: dithiothreitol, 124 Pro-Lab Diagnostics, UK) prior to dilution in either NFW or 10% Chelex® 100 Resin Laboratories, Watford, UK) 23 . Mucolyse TM was also added 1:1 to the final pool of negative saliva 126 samples and the second SARS-CoV-2 spiked pool. 127 128 RNA extraction 129 purification kit (Thermofisher). Briefly, 10 µl of sample (diluted in 190 µl DEPC treated water) was 139 added to 700 µl of prepared lysis solution. Samples were then inactivated for 10 minutes at room 140 temperature within the safety cabinet before automated RNA extraction using a KingfisherFlex 141 (Thermofisher). RNA was eluted in 90 µl of NFW. 142 143 144 All rRT-PCR assays were performed in single replicates using 5 µl of RNA template. The saliva sample 145 collected within the HHFT was analysed using the COVID-19 genesig® Real-Time PCR assay 146 (Primerdesign Ltd, Chandler's Ford, UK) according to the manufacturer's guidelines, on a MIC qPCR 147

Cycler (Bio Molecular Systems, London, UK). The cycling conditions were adjusted to the following: a 148 reverse-transcription (RT) step of 10 minutes at 55 o C, a hot-start step of 2 minutes at 95 o C, and then 149 45 cycles of 95 o C for 10 seconds and 60 o C for 30 seconds. The Genesig® COVID-19 positive control 150 included in the kit, a negative extraction control, and a no template control were also included on 151 each rRT-PCR run. 152

The saliva samples collected from the UHS and the spiked whole virus dilution series were tested using 154 the E gene RT-PCR as described previously (Corman et al.,2020) using the AgPath-ID™ PCR kit 155 (Thermofisher). Samples were run on an Aria qPCR Cycler (Agilent) and results analysed using the 156 Agilent AriaMX 1.5 software. The cycling conditions were adjusted to the following: a reverse-157 transcription (RT) step of 10 minutes at 55 o C, a hot-start step of 3 minutes at 94 o C, and then 45 cycles 158 of 94 o C for 15 seconds and 60 o C for 15-30 seconds during data acquisition. The SARS-CoV2 positive 159 control RNA, a negative extraction control, and a no template control were also included on each rRT-160 PCR run. 161 162 163 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted July 21, 2020. . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 21, 2020. . https://doi.org/10. 1101 /2020 in 640, with and without heat treatment (70 o C for 4 minutes or 98 o C for 2 mins) was added to the 191 reaction. Heating was performed on a dry heat block. The same treatments were applied to the saliva 192 pools spiked with whole inactivated virus and to the non-spiked negative saliva pool, however the first 193 spiked saliva pool was only titrated as far as 1 in 40 in the first instance. After addition to Direct LAMP all treatments were pooled according to dilution (e.g. all temperature treatments were pooled 195 according to the dilution) and extracted for rRT-PCR analysis (for the first spiked sample [ 

Optimization of the Direct RT-LAMP assay for detection of SARS-CoV-2 was determined using three 201 positive saliva samples, a pool of non-spiked negative saliva and a pool of spiked saliva. 202

The three positive saliva samples diluted 1:1 in Mucolyse TM rRT-PCR CT values were 21.08, 24.47 and 203 25.27 (Table 1) . For the first spiked saliva pool, the whole inactivated virus spiked into saliva prior to 204 dilution rRT-PCR CT was 26.70 (Table 2) . For the second spiked saliva pool, the whole inactivated virus 205 spiked into saliva diluted 1:1 in Mucolyse TM (prior to further dilutions) rRT-PCR CT was 22.86 (Table 3) . 206 207 From rRT-PCR data samples were assessed for sensitivity using the Direct-LAMP protocol. Samples 208

were assessed in order of highest viral load by rRT-PCR (CT 21.08: The saliva sample with the highest viral load (CT 21.08) when diluted in water was detected in 212 duplicate in five dilutions (1 in 40 to 1 in 640) without heat treatment, in all eight dilutions (1 in 5 to 1 213 in 640) following 70 o C for 4 mins and in seven dilutions (1 in 5 to 1 in 320) following 98 o C for 2 mins 214 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 21, 2020. . https://doi.org/10.1101/2020.07.16.20155168 doi: medRxiv preprint detected in duplicate in seven dilutions (1 in 10 to 1 in 640) without heat treatment and in all eight 216 dilutions (1 in 5 to 1 in 640) following either 70 o C for 4 minutes or 98 o C for 2 minutes (Table 1, (Table 2) . When 239 diluted in 10% (w/v) Chelex® 100 Resin the same saliva sample was detected in duplicate in two 240 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 21, 2020. . https://doi.org/10. 1101 /2020 following 70 o C for 4 minutes and in all four dilutions (1 in 5 to 1 in 40) following 98 o C for 2 minutes 242 (Table 2) . 243 244 A further dilution series of SARS-CoV-2 inactivated whole virus was prepared to include the 1:1 245

Mucolyse TM dilution that is used for clinical samples and to extend beyond a 1 in 40 dilution to reach 246 the limit of detection of the Direct RT-LAMP assay. The whole inactivated virus spiked into saliva (CT 247 of 22.86 when diluted in water was detected in duplicate at one dilution (1 in 80) without heat, in 248 three dilutions (1 in 5, 1 in 10 and 1 in 80) following 70 o C for 4 mins and in three dilutions (1 in 5, 1 in 249 10 and 1 in 40) following 98 o C for 2 mins (Table 3) . When diluted in 10% (w/v) Chelex® 100 Resin the 250 same saliva sample was detected in duplicate in three dilutions (1 in 20, 1 in 40 and 1 in 80) without 251 heat treatment, in six dilutions (1 in 5 to 1 in 160) following 70 o C for 4 minutes and in six dilutions (1 252 in 5 to 1 in 160) following 98 o C for 2 minutes (Table 3) . 253

Discussion 255

This study describes the rapid optimization of a method to permit direct detection of SARS-CoV-2 RNA 256 within saliva samples using RT-LAMP, without need for prior RNA extraction. Our previous publication 257 was focused on optimizing conditions for rapid detection of SARS-CoV-2 within viral transport media 258 from swabs samples 22 . In that publication, preliminary evaluation of the direct transfer of the swab 259 sample preparation method for comparable detection in paired saliva samples was poor, indicating 260 that a different sample preparation method would be required for optimal detection of SARS-CoV-2 261 RNA in crude saliva. In this study we show for the first time that the optimal sample preparation 262 method to allow SARS-CoV-2 RNA detection within crude saliva samples (1:1 mix of saliva and 263 . CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 21, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 21, 2020. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 21, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 21, 2020. CC-BY-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted July 21, 2020 . . https://doi.org/10.1101 /2020 

Dark grey shading indicates samples positive in duplicates by direct RT-LAMP, Light grey shading indicates samples positive in single replicates. Blank wells represent no amplification detected (negative) in Direct RT-LAMP