key: cord-0742982-6uci7nu4
authors: Mohon, A. N.; Hundt, J.; van Marle, G.; Pabbaraju, K.; Berenger, B.; Griener, T.; Lisboa, L.; Church, D.; Czub, M.; Greninger, A.; Jerome, K.; Doolan, C.; Pillai, D. R.
title: Development and validation of direct RT-LAMP for SARS-CoV-2
date: 2020-05-04
journal: nan
DOI: 10.1101/2020.04.29.20075747
sha: 39b8efc748bbaef79d9e450a666da5c29a9f6746
doc_id: 742982
cord_uid: 6uci7nu4

We have developed a reverse-transcriptase loop mediated amplification (RT-LAMP) method targeting genes encoding the Spike (S) protein and RNA-dependent RNA polymerase (RdRP) of SARS-CoV-2. The LAMP assay achieves the same limit of detection as commonly used RT-PCR protocols based on artificial targets, recombinant Sindbis virus, and clinical samples. Clinical validation of single target LAMP (N=108) showed a positive percent agreement (PPA) of 33/34 (97.1%) and negative percent agreement (NPA) of 73/74 (98.6%) compared to reference RT-PCR. Dual target RT-LAMP achieved a PPA of 11/11 (100%) and NPA 13/13 (100%) when including discrepant samples. The assay can be performed without a formal extraction procedure, with lyophilized reagents that do need cold chain, and is amenable to point-of-care application with visual detection.

and North America (5). A common theme in the public health response to COVID19 and similar threats is 40 the lack of rapidly deployable testing in the field to screen large numbers of individuals in exposed areas, 41 international ports of entry, and testing in quarantine locations such as the home residences, as well as 42 low-resourced areas (6). This hampers case finding and increases the number of individuals at risk of 43 exposure and infection. With the ease of travel across continents, delayed testing and lack of screening 44 programs in the field, global human-to-human transmission will continue at high rates. These factors 45 make a pandemic very difficult to contain. Early identification of the virus and rapid deployment of a 46 targeted point of care test (POCT) can stem the spread through immediate quarantine of infected 47 persons(7). We used existing viral genome sequences to develop a SARS-CoV-2 loop mediated 48 amplification (LAMP) assay for clinical use and evaluated whether an extraction-free and instrument-49 free approach could be achieved (8, 9) . POCT requires portability and low complexity without reliance on 50 sophisticated extraction and read-out instrumentation. Furthermore, LAMP relies on an alternate set of 51 reagent chemistry that does not depend on or hinder critical elements of the RT-PCR supply chain which 52

Four fragments of specific SARS-Co-V2 regions (ORF1ab (nsp 3,10-11), RdRP (nsp 12), and spike (S)) were 76 synthesized by SGI-DNA Inc. (San Diego, CA) . Fragments were ligated to make one large concatenated 77 DNA template using the BioXP3200 (SGI-DNA, San Diego, CA) automated Gibson assembly system. The 78 final template was 1097 base pairs long containing concatenated ORF1a/b (nsp 3)-Spike protein-RdRP 79 (nsp 12) -ORF1a/b (nsp 10-11) fragments together with flanking plasmid sequence in that order. 80

The RT-PCR assays used in this study were performed according to previous publications for the E 82 gene(15) and N2 gene (16). The E gene RT-PCR was performed with modification according to the Alberta 83

Public Health Laboratory reference method (17). Five (5) L of input template was used to perform the 84 RT-PCR reactions on the same day when RT-LAMP was performed. A maximal Ct value cut-off of 40 was 85 used to determine positivity for all RT-PCR reactions. 86

The single-target LAMP reaction was conducted using a combination of Warmstart Rtx Reverse 88 CC-BY-NC-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 May 4, 2020 . . https://doi.org/10.1101 or fluorescence values on the QuantStudio 5 RT system (ThermoFisher, 98 Toronto, ON). RFU values above 100 on the CFX-96 instrument or greater than 100,000 on the 99 QuantStudio 5 instrument were considered positive when associated with a typical amplification curve 100 before a 30-minute reaction time. Each primer set was used to amplify the artificial viral DNA target at 101 different temperature (61°C, 63°C, and 65°C) for a maximum LAMP assay run time of 40 minutes. 102

A positive LAMP assay was detected visually by the pre-addition of 0.5µL colorimetric fluorescence 104 indicator (CFI). CFI was made up of the combination of 0.7% (v/v) 10000X Gelgreen (Biotium, Freemont, 105 CA) in 12 mM Hydroxynapthol blue resuspended in dH20 (Sigma-Aldrich, Oakville, ON). After the 30-106 minute LAMP reaction time, the tubes were exposed to blue LED light using a Blue Light Transilluminator 107 (New England Biogroup, Atkinson, NH) to visualize the green fluorescence. 108

In order to determine if the LAMP primers and master mix could be lyophilized, oligonucleotides were 110 shipped to Pro-Lab Diagnostics (Toronto, Canada) and lyophilized with GspSSD2 isothermal master 111 mixture (Optigene, UK). The lyophilized primer, enzyme, master mix combination was hydrated in 15 L 112 of resuspension buffer (Pro-Lab Diagnostics) to which 10 L of the sample was added. Both direct LAMP 113 (see method described later) and kit-based RNA extractions were performed in this way. 114

Limit of detection of the LAMP assay was evaluated in three different ways. Initially, copy number of the 116 synthesized DNA fragment was determined by comparing concentration and molecular weight. 117

Subsequently, the template solution was serially diluted to achieve a range from 5,000,000 to 5 copies 118 per reaction. These serially diluted templates were tested by all primers sets. Secondly, the extracted 119 . CC-BY-NC-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 May 4, 2020. . https://doi.org/10. 1101 RNA from one patient specimen was 10-fold serially diluted and tested with the LAMP assay and two RT-120 PCR assays used by reference laboratories targeting both the E gene(15) and N2 gene(16). Finally, the 121 artificial template containing the targeted sequences of interest was cloned into Sindbis Virus (SV) viral 122 vector system (SINrep5) containing green fluorescent protein (EGFP) and then transfected into BHK-21 123 cell lines (18) source. All samples in the set were tested simultaneously using the E gene RT-PCR described above and 140 S gene LAMP (primer set 2). Discrepant analysis was performed by performing the CDC N2 gene(16) RT-141 PCR assay. For the LAMP assay, 10 L of the RNA extract was used for each reaction. A second validation 142 . CC-BY-NC-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 May 4, 2020. . https://doi.org/10. 1101 study was performed for the dual-target LAMP (S gene and RdRP). Here, RT-PCR-positive clinical (n=34), 143 contrived (n=10), and 72 negative NP swab samples were tested. All LAMP reactions for the clinical 144 validation studies was performed using the combination of Warmstart Rtx Reverse Transcriptase with 145 Bst 2.0 Warmstart DNA Polymerase as described earlier. 146

Direct LAMP assay without formal extraction 147 A direct LAMP assay was also conducted to establish an extraction-free approach. In this scheme, the 148 LAMP reaction mixture was prepared without the enzymes. 9.5 µL reaction mixture containing all 149 reagents except the enzymes was added to 14 µL of the 10-fold serially diluted NP VTM sample. For 150 direct LAMP, VTM must be diluted 1:10 (v/v) with dH20 prior to addition. The mixture was then heated 151 at 95°C for 3, 5, and 10 minutes to both inactivate virus and presumptively release viral RNA. Finally, Bst 152 2.0 WarmStart ® DNA Polymerase (1 µL) and Warm Start® reverse transcriptase (0.5 µL) were directly 153 added to the reaction mixture and the LAMP assay was carried out as above. Due to evaporative loss, 154 boil steps should have excess volume to ensure adequate input template for LAMP reactions. 155

In silico analysis of primer combinations to determine cross-reactivity 156 A blast search alignment (https://blast.ncbi.nlm.nih.gov/Blast.cgi) for primers in set 2 (spike gene) and 157 set 3 (RdRP gene) were performed against a critical list of infectious agents that cause upper respiratory 158 tract infections. A nucleotide local alignment using BLASTn with the default parameters was performed 159 against the National Center of Biotechnology Information (NCBI) Nucleotide database. 160

In order to determine the limits of detection of the five LAMP primer sets designed for SARS-CoV-2, 163 experiments were conducted using an artificial gene target construct. Three targets ( Figure 1 ) selected 164 . CC-BY-NC-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 May 4, 2020. . https://doi.org/10. 1101 were the Spike (S) protein gene, the RNA-dependent RNA polymerase (RdRP) region (nsp 12), and nsp 3 165

region in the open reading frame (ORF) 1a/b protein encoding gene sequence. Figure 2 shows the 166 amplification curves for the five primer sets used in this study. Based on time to positivity, primer sets 1, 167 2, and 3 demonstrated the fastest positive reaction time, suggesting optimal performance in the LAMP 168 assay at 50 copies of the artificial target per reaction. When tested at 5 copies per reaction, primer set 2 169 targeting the S gene showed the best limit of detection. An in silico analysis of primer set 2 (S gene) and 170

set 3 (RdRP) demonstrated no significant sequence alignment cross-reactivity with known upper 171 respiratory tract infectious pathogens (data not shown). 172

An advantage of LAMP is the ability to detect amplification with the naked eye either via the use of 174 colorimetric or fluorescent dyes. To test this, LAMP was conducted using the S gene LAMP primer set 175 using the artificial gene target. Figure 3 shows a positive test based on fluorescent detection using a blue 176 light emitting diode (LED). A serial dilution between 50 x 10 7 and 5 copies per reaction of the artificial 177 gene construct is shown with a limit of detection of 50 copies per reaction obtained. The LAMP assay 178 using enzyme GspSSD2 was also performed using a lyophilized master mix. Studies were performed with 179 both kit-based RNA extracted clinical samples as well as using the direct LAMP method described later. 180

These data showed that amplification occurred up to a 1000-fold dilution of a clinical sample with the 181 direct LAMP method (data not shown). 182

In order to determine LOD based on viral titer, recombinant RNA viral vector from Sindbis virus (SV) 184 containing SARS-CoV-2 targets was generated. LAMP was performed using Set 2 (S gene) alone and in 185 combination with Set 3 (RdRP region (nsp 12)). The primer sets in question achieved a LOD of 0.1-1.0 SV 186 EGFP forming units per L for the S gene alone and the dual-target LAMP amplifying both S and RdRP 187 . CC-BY-NC-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 May 4, 2020. . https://doi.org/10. 1101 genes in a single reaction (Table 2) . One SV EGFP focus forming unit (FFU) roughly corresponds to 1.0 188 infectious viral particle given the assumption that one virus particle infects one cell. 189

Direct LAMP detection of SARS-CoV-2 without formal extraction 190

Given extraction reagent supply chain shortages, we tested the ability of LAMP using the S gene to 191 amplify a SARS-CoV-2 positive nasopharyngeal swab specimen without a formal extraction procedure. 192

The VTM was heated at 95 o C for 3, 5 and 10 minutes and then subjected to the S gene LAMP procedure 193 in a serial dilution experiment. This experiment was performed in a head-to-head comparison with the 194 same specimen tested after formal extraction. Figure 4 demonstrates that amplification occurred with 195 direct LAMP up to a dilution of 100,000-fold with a 95 o C for 3 minutes heat step in a single experiment. 196 Table 3 shows the data for direct LAMP compared to LAMP and RT-PCR from RNA extracts in triplicate 197 experiments. Reproducible amplification with direct LAMP (95 o C for 3 minutes) occurred at a dilution 198 factor of 1,000-fold, whereas LAMP from an RNA extract was successful reproducibly at a dilution of 199 100,000-fold. The same experiment was conducted comparing E gene RT-PCR to RT-LAMP (dual-target S and RdRP 207 gene) as well as a direct RT-LAMP (dual-target S and RdRP gene) without formal extraction (boil method) 208 on a separate specimen. These data are shown in Figure 5 and Table 4 . Reproducible amplification 209 . CC-BY-NC-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 May 4, 2020. . https://doi.org/10. 1101 occurred at 1000-fold dilution for RT-PCR (E gene) and RT-LAMP. Direct-LAMP amplified reproducibly at 210 a 100-fold dilution for this sample. 211

Clinical validation of SARS-CoV-2 LAMP using nasopharyngeal swab samples 212 A sample set of nasopharyngeal swabs (n=108) from COVID-positive, COVID-negative, together with 213 samples for other respiratory viruses were used to validate the S gene LAMP primer set (Table 5) represented a true positive that may be due to sample decay in storage. The second falsely negative 222 sample was positive by all methods including the RdRP LAMP primer set and was deemed a false 223 negative in the final analysis. In order to eliminate the concern for S gene false negatives, dual-target S 224 and RdRP LAMP was performed to increase clinical sensitivity. This second clinical sample set 225 intentionally included low positive (high Ct value) specimens. CC-BY-NC-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 May 4, 2020. . https://doi.org/10.1101/2020.04.29.20075747 doi: medRxiv preprint

The global pandemic with SARS-CoV-2 has resulted in the need for diagnostic test development at a 234 scale never seen before. Rapid deployment of validated laboratory-developed diagnostic tests or 235 commercial tests was essential to the containment of the virus as it allows for self-quarantine measures 236 to be imposed in a strategic fashion before widespread community transmission occurs (6, 7) . 237

Diagnostic tests have to be analytically sensitive in order to not to miss any cases in the acute phase of 238 viremia (11). As such, NAATs serve this purpose. In particular, RT-PCR has been employed as the primary 239 diagnostic counter-measure (20). However, reagent supply chains for key items are under immense 240 pressure. Local solutions to reagent sources have become paramount because barriers to trade of these 241 selected items have been a concern. 242 We noted that a one false negative occurred with single-target LAMP (S gene) for a lower Ct value 243 sample that may be due to genetic polymorphism at key residues where LAMP primers bind. This issue 244 was overcome by using two targets (S and RdRP genes) simultaneously. The single and dua-target RT-245 LAMP test for SARS-CoV-2 has comparable analytical sensitivity and achieved excellent agreement with 246 the reference method. We noted that at high Ct value (>35), and presumably low-level infection, E gene 247 RT-PCR identified specimens that LAMP did not. This suggests the LOD of the RT-PCR used in this study 248 may be superior to LAMP for these low positives. However, we note that samples with E gene RT-PCR Ct 249 values greater than 35 are relatively rare (~1%) at our reference laboratory (our unpublished 250 observations). Increasing LAMP reaction times may reduce false negatives, but also lead to spurious 251 amplification. The clinical relevance of these low positives is still not well understood and could 252 represent early infection during the incubation period, late infection after the initial viral peak, or 253 asymptomatic carriage. The transmissibility of these low positives to others is also not well understood. 254

Moreover, if we assume an overall prevalence of 5%, the negative predictive value of LAMP is 99.56% 255 . CC-BY-NC-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 May 4, 2020. . https://doi.org/10.1101/2020.04.29.20075747 doi: medRxiv preprint (95% CI 98.89 -99.83) and positive predictive value is 100%. These values clearly support the use of 256 LAMP as an alternative NAAT. 257 LAMP does not rely on the same reagents as RT-PCR and thus alleviates pressure on key supply chain 258 items. The LAMP method is amenable to high throughput testing in either 96-well or 384-well. The assay 259 is also able to detect SARS-CoV-2 in VTM without the need for a kit-based RNA extraction method 260 relying on commercial reagents. However, we noted a 2-log drop in analytical sensitivity when direct 261 LAMP was performed following this heat step. The drop in analytical sensitivity will only affect low-level 262 viral load specimens. This is still within a good range compared to RT-PCR using RNA isolation. We 263 believe that addition of a buffer to stabilize the RNA-enzyme complex in the LAMP reaction may further 264 enhance this extraction-free approach. This needs to be tested. The heat step at 95 o C for 3 minutes 265 should significantly inactivate the virus permitting safe operation of the assay in a Class II biosafety 266 cabinet (21). Taken together, these data support the use of LAMP chemistry as an alternate method for 267 laboratory developed NAATs. 268

Our studies with a LAMP enzyme called GspSSD2 also provided encouraging results. These data 269 demonstrated that lyophilized GspSSD2 and reagents are able to amplify SARS-CoV-2 directly from a 270 specimen without a kit-based RNA extraction. Additionally, visual detection with a simple blue LED light 271 is able to discriminate positive from negative. These features are particularly useful for resource-limited 272 settings without sophisticated laboratory infrastructure or where the cost of or access to kit-based 273 reagents and equipment are prohibitive. Further studies are required to clinically validate this low-cost 274 approach. 275

Limitations of the study include not testing other sample types such as alternate swabs, nasal washes, 276 oropharyngeal samples, sputum, or stool. This work is ongoing with a special emphasis on swab-free 277 testing. Also formal SARS-CoV-2 viral titers were not calculated in the limit of detection studies. 278

. CC-BY-NC-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 May 4, 2020 . . https://doi.org/10.1101 Nevertheless, LAMP presents a much needed alternative approach to SARS-CoV-2 diagnostic testing that 279 is available for deployment immediately in a LDT format as it relies on other key reagents that do not 280 cannibalize RT-PCR reagents. Ultimately, the aim is to port LAMP chemistry on a stand-alone microfluidic 281 device POCT to be deployed in the community, either at ports of entry, homes, pharmacies, or resource-282 limited settings. 283

Dr. Ranmalee Amarasekara for expert technical assistance, Daniel Castaneda Mogollon for performing 285 bioinformatics analysis, and Omar Abdullah for research analytical support. CC-BY-NC-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 May 4, 2020. CC-BY-NC-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 May 4, 2020. CC-BY-NC-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 May 4, 2020. . https://doi.org/10.1101/2020.04.29.20075747 doi: medRxiv preprint CC-BY-NC-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 May 4, 2020 . . https://doi.org/10.1101 CC-BY-NC-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 May 4, 2020 . . https://doi.org/10.1101 . CC-BY-NC-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 May 4, 2020 . . https://doi.org/10.1101 CC-BY-NC-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 May 4, 2020 . . https://doi.org/10.1101 Tables: 479   Table 1 : Primer sets used in this study to perform RT-LAMP. All 5 primer sets are shown: Set 1 (ORF1a/b, 480 nsp3); Set 2 (S gene); and Set 3, 4, and 5 (RdRP). 481 482 Primer name Sequence S1-F3 AGTTTGAGCCATCAACTCA S1-B3: TGAACCTCAACAATTGTTTGA S1-F1P CAGGTTGAAGAGCAGCAGAAGTGTACTGAAGATGATTACCAAGG S1-B1P AGCAAGAAGAAGATTGGTTAGATGATGTCTGATTGTCCTCACTG S1-LPF GGCACCAAATTCCAAAGGT S1-LPB AACTGTTGGTCAACAAGACGG 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 May 4, 2020. . CC-BY-NC-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 May 4, 2020. . Direct dual RT-LAMP 3 min/ 95 o C 10 6/6 9/9 9/9 100 6/6 9/9 9/9 1000 6/6 5/9 1/9 10000 1/6 1/9 0/9 100000 0/6 0/9 0/9 . CC-BY-NC-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 May 4, 2020. . https://doi.org/10. 1101 . CC-BY-NC-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 May 4, 2020. . https://doi.org/10. 1101 

Avian flu, SARS, MERS, Ebola, Zika… what next? 300 Vaccine

Early Transmission Dynamics in Wuhan, China, of Novel 305 Coronavirus-Infected Pneumonia

A pneumonia outbreak associated with a new 309 coronavirus of probable bat origin

A new coronavirus 312 associated with human respiratory disease in China

Severe Outcomes Among Patients with Coronavirus 314 Disease 2019 (COVID-19) -United States

Diagnostic Testing for Severe Acute Respiratory Syndrome-Related Coronavirus-2: A Narrative 318 Review

Novel Coronavirus Disease (COVID-19): Paving the 320 Road for Rapid Detection and Point-of-Care Diagnostics

Phylogenetic network analysis of SARS-CoV-2 322 genomes

Genomic characterisation and 326 epidemiology of 2019 novel coronavirus: implications for virus origins and receptor binding

Predicting the impacts of epidemic outbreaks on global supply chains: A 329 simulation-based analysis on the coronavirus outbreak (COVID-19/SARS-CoV-2) case

Risk prediction for 333 severe disease and better diagnostic accuracy in early dengue infection; the Colombo dengue study

Ultrasensitive loop mediated isothermal 336 amplification (US-LAMP) to detect malaria for elimination

Prevalence and epidemiological characteristics of asymptomatic malaria based 339 on ultrasensitive diagnostics: A cross-sectional study

Molecular Diagnostic Field Test for Point-of-Care Detection of Ebola Virus 342 Directly From Blood

Drosten 346 C. 2020. Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR