key: cord-329687-vhi4tbnc
authors: Verdugo, C.; Plaza, A.; Acosta-Jamett, G.; Castro, N.; Gutierrez, J.; Hernandez, C.; Lopez-Joven, C.; Loncoman, C.; Navarrete, C.; Ramirez-Reveco, A.; Romero, A.; Silva, A.; Vega, M.; Vergara, J.
title: A comparative evaluation of dye-based and probe-based RT-qPCR assay for the screening of SARS-CoV-2 using individual and pooled-sample testing.
date: 2020-06-03
journal: nan
DOI: 10.1101/2020.05.30.20117721
sha: 
doc_id: 329687
cord_uid: vhi4tbnc

Effective interventions are mandatory to control the transmission and spread of SARS-CoV-2, a highly contagious virus causing devastating effects worldwide. Cost-effective approaches are pivotal tools required to increase the detection rates and escalate further in massive surveillance programs, especially in countries with limited resources that most of the efforts have focused on symptomatic cases only. Here, we compared the performance of the RT-qPCR using an intercalating dye with the probe-based assay. Then, we tested and compared these two RT-qPCR chemistries in different pooling systems: after RNA extraction (post-RNA extraction) and before RNA extraction (pre-RNA extraction) optimizing by pool size and template volume. We evaluated these approaches in 610 clinical samples. Our results show that the dye-based technique has a high analytical sensitivity similar to the probe-based detection assay used worldwide. Further, this assay may also be applicable in testing by pool systems post-RNA extraction up to 20 samples. However, the most efficient system for massive surveillance, the pre-RNA extraction pooling approach, was obtained with the probe-based assay in test up to 10 samples adding 13.5 uL of RNA template. The low cost and the potential use in pre-RNA extraction pool systems, place of this assays as a valuable resource for scalable sampling to larger populations. Implementing a pool system for population sampling results in an important savings of laboratory resources and time, which are two key factors during an epidemic outbreak. Using the pooling approaches evaluated here, we are confident that it can be used as a valid alternative assay for the detection of SARS-CoV-2 in human samples.

Two segments of the nucleocapsid gene (i.e. N1 and N2) were amplified using 4x TaqMan® Fast Virus 1-step Master Mix (Thermo Fisher) in a 20 µL reaction containing 5 µL of master mix, 5 µL of RNA template, 1.5 µL of primer and probe mix (2019-nCOV CDC EUA Kit, IDT), and 8.5 µL of nuclease-free water. Briefly, the cycling profile was 25ºC for 2 min, 50ºC for 15 min, and 95ºC for 2 min, followed by 45 cycles of 95ºC for 3 sec and 55ºC for 30 sec. Adequate internal controls, viral positive control and negative control were included during each run.

The analytical sensitivity (i.e. minimum number of copies in a sample measured accurately), expressed by the limit of detection (LOD) of the assay was determined using six 10-fold serial dilutions of a synthetic plasmid of the full nucleocapsid gene according to datasheet information (2019-CoV Plasmid Control, IDT), from 1x10 5 to 1 viral copies/µL in ultra-pure water. Dilutions were run in duplicate through both, dye-based and probe-based, assays. A standard curve was obtained plotting the CT values against the copy number. The reaction efficiency was determined by the calculation of the correlation coefficient (R 2 ). The repeatability and reproducibility of the assay was evaluated in four 10-fold serial dilutions of the standard plasmids (from 1x10 5 to 1x10 2 ) in duplicates using intra-and inter-assay tests within the same run and through two independent runs, respectively. The mean, standard deviations (SD), and coefficient of variation (CV) were calculated for each dilution based on their CT values.

We compared the performance of the dye-based with the probe-based assay by analyzing 610 nasopharyngeal samples. Viral RNA was isolated as describe above. Each sample was individually analyzed using both methods in parallel, using adequate internal controls, viral positive control and negative control.

In order to evaluate the ability of the RT-qPCR assay to scale for massive use, we run our tests in two different group systems: pooling after RNA extraction (post-RNA extraction) and pooling before RNA extraction (pre-RNA extraction) ( Figure 3B and 3C). For the pooling post-RNA extraction, the eluted RNA from a confirmed positive sample of a patient was added to pools of 5, 10, 15 and 20 RNA negative samples. The positive sample was added at two different virus copy numbers emulating infected individuals at mid (~400 copies/µL) and a high (~6,000 copies/ µL) viral load (Kim et al 2020; Zou et al 2020) . A total of 5 µL of RNA sample was added to the final RT-qPCR mix. Each pool was run in duplicate. An only negative pool were run within each . 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 June 3, 2020. . https: //doi.org/10.1101 //doi.org/10. /2020 assay. A positive plasmid control (PC) at 2,000 viral copies/µl was included. For the pooling pre-RNA extraction, we tested pools of 5 and 10 samples using both RT-qPCR assays. A 100 µL of viral transport media of each individual samples were mixed into a single tube. Then, a total of 200 µL of the mix was used for viral extraction, as described above, with a final elution of 60 µL.

We compared adding volumes of 5 µL and 10 µL of RNA template added to the RT-qPCR mix.

We evaluated 10 pools using a set of positive samples with CT values from 14.76 to 38.2, emulating patients from a wide spectrum of viral load. We further compared 15 pools of 10 samples each adding 5 µL and 13.5 µL of RNA template, which is the maximum volumen allowed in a 20 µL RT-qPCR mix. All pools include one positive sample that with a CT values that ranged from the 12.7 to 39.9. An only negative pool was run within each assay.

Evaluation of a dye-based assay.

All set of primers targeting nucleocapsid gene amplified correctly using the dye-based real-time RT-PCR assay. One pair of primers (N1-F and N1-R) was selected for further efficiency and performance based on the highest sensitivity (lowest CT) obtained minimizing non-specific amplification products during melting curve analysis. The standard curve of the CT against the amount of synthetic viral plasmid showed a wide dynamic range from 1 to 1x10 5 viral copies/µL in 40 cycles, a correlation coefficient, R 2 , of 0.99 with a standard curve slope of -3.715, and a reaction efficiency, E, of 86% ( Figure 1A ). The limit of detection of SARS-CoV-2 using the dyebased assay was at a dilution of 10 viral copies/µL with a mean CT value of 35.31. This value was 1.09 cycles higher than the probe-based assay at the same dilution ( Figure 1B) . Further, across all dilutions, the CT difference among assays was almost zero (i.e. 0.08), ranging from -1.09 to 0.87 cycles ( Table 2 ). The dissociation curve performed after completed RT-qPCR peaked at 83.51 ± 0.25 ºC ( Figure 2 ). All internal controls amplified correctly. No positive fluorescence amplification signal was obtained from negative controls and nuclease-free water.

Standard curves for each assay were highly reproducible with no significant differences in slopes among runs of the same assay. The intra-assay SD and CV ranged from 0.02 to 0.14 and 0.1 to 0.44%, respectively, whereas the inter-assay ranged from 0.04 to 0.18 and 0.05 to 0.43% (Table 3 ), indicating that the dye-based assay is highly repeatable and reproducible.

Performance comparison among assays A total of 610 nasopharyngeal samples were examined by dye-and probe-based assays in order to compare performance among assays. The probe-and dye-based RT-qPCR detected 24 and 20 positive samples to SARS-CoV-2, respectively. The estimated sensitivity and . 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 June 3, 2020. . https://doi.org/10.1101/2020.05.30.20117721 doi: medRxiv preprint specificity of the dye-based RT-qPCR was 83.3% (95% CI: 62.9-95.3%) and 100% (95% CI: 99.4-100%), respectively. In addition, positive samples from probe-based showed always a lower CT value than dye-based assays (range: 0.72 to 5.9 cycles).

We compared the RT-qPCR performance of individual samples ( Figure 3A ) with pooled samples grouped at two different analytical stages: after ( Figure 3B ) and before ( Figure 3C ) the viral RNA extraction. In the post-RNA extraction pools, RNA of SARS-CoV-2 was detected in all pools, independent the amount of starting viral material, emulating patients with two different viral loads on each pool. All pools, in both dye-and probe-based, showed a negative impact on the performance of the RT-qPCR, increasing the CT values when compared with individual testing ( Figure 4 ). Using a high viral copies sample (~6,000 copies/µL), the mean CT values of pools ranged from 27.14 to 29.23 that were 4.12 to 6.22 cycles from the original sample using the dye-based assay, and from 23.84 to 25.68 at 2.59 to 4.43 cycles away from the original sample using the probe-based sample (Table 4) . Similarly, pools emulating a patient with a mid-viral load (~400 copies/µL) showed CT values from 30.40 to 31.83 which were 3.10 to 4.53 cycles from the original sample in the dye-based assays, whereas using the probe-based assay the CT values were less than 2 cycles from the original sample (Table 4 ). None of the negative pools by probe-based RT-qPCR showed fluorescence amplification signal. The same negative pools showed a high CT value using the dye-based assay, although no melting products were detected at the 83°C target peak.

In the pre-RNA extraction pools, we evaluated pools using individual positive samples with a wide range of individual CT values. The probe-based assay was able to detect viral amplification in 8/10 pools of size 5 and 10, using either 5 or 10 µL of RNA template. In all pools, the CT values increased when compared with individual testing. As expected, the CT was greater when the pool size increased from 5 (CT mean difference of 2.83) to 10 (CT mean difference of 7.1) (Table 5 ). When the amount of RT-qPCR template was doubled from 5 to 10 µL, the CT difference with the individual sample decreased ( Figure 5A and B). We were able to detect pools that contained an individual positive sample with CT value up to 35. The two pools containing samples with the highest individual CT values, 37.96 and 38.2, were not detected by pooling in either 5 or 10 and adding 5 or 10 µL. When the volume of RNA template increased from 5 to 13.5 µL, all pools of 10 samples (15/15) containing individual positive samples with CT values from 12.7 to 39.9 were detected. In contrast, 13/15 pools adding 5 µL of RNA template were detected, where two pools with individual positive sample CT of 33.5 and 39.9 were not detected (Table 6 ). All CT values decreased when the volume of RNA template increased ( Figure 5C ).

. 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 June 3, 2020. . https://doi.org/10. 1101 /2020 The performance of the dye-based assay in pre-RNA extraction pools resulted in the amplification of 6/11 and 5/11 in pools of size 5 and 10, respectively, using 5 µL of RNA template (Table 6) . Surprisingly, increasing the amount of RNA template to 10 µL did not increase the detection rate. In fact, adding more template to the RT-qPCR mix, the detection decrease in 3/11 and 0/11 in pools of 5 and 10, respectively,

In this study, we compared a dye-based and probe-based real-time RT-qPCR assay for the economic and rapid detection and quantification of SARS-CoV-2 in human samples by individual and pooling testing. The dye-based technique showed a high analytical sensitivity similar to the probe-based detection assay used worldwide. Further, we showed that this assay may also be applicable in testing by pool systems post-RNA extraction, whereas the probebased assay have a excellent performance in pre-RNA extraction pooling tests. Thus, each test has particular advantages that can be exploited for massive use to increase sensitivity in surveillance programs in large target populations.

Our results showed that using the dye-based assay describe here, the SARS-CoV-2 can be detected up to 50 viral copies (a dilution of 10 copies/µL) of the RNA template, showing a similar analytical sensitivity obtained with the probe-based standard technique ( Table 2) showed no significant differences in CT values between pooled and individual samples, suggesting that sensitivity was not affected by pooling specimens, regardless of the viral load (Wacharapluesadee et al, 2020). We further explored the use of pooling samples previous to the analytical stage of RNA extraction. This is intended to significantly save time, laboratory staff effort, and economic resources in a surveillance program. Using the probe-base assay, we were able to detect positive individuals with CT values of 31.09 in pools of 5 and 10 samples using 5 or 10 µL of RNA template. Pooled samples with one individual with CT over 38 were not detected. Due to the lack of samples between CT 32 and 38, we cannot rule out the analytical sensitivity of this assay at CT 32 using 10 µL of RNA template but further analyses might be pursued to detect a more trustful detection threshold. Since increasing the amount of the RNA template from 5 to 10 µL we observed less compromise of cycles, we further explore increasing to the maximum amount of RNA template allowed in a 20 µL RT-qPCR mix, with a set of positive samples with a wider range of CT values (12.7 to 39.9). When the template volume increased to 13.5 µL in pools of 10, we were able to reduce the delta CT from 6.25 (using 5 µL of RNA template) to 3.72 on average and increase the detection rate to 100%.

Contrary to our expectations, we observed an increase in the CT value the RNA template volume increased using the dye-based assay. The increase in CT value when using 10 µL versus 5 µL of RNA template in the dye-based RT-qPCR assay could be explained by the effect of inhibitors present in the extracted RNA sample, although this effect was not observed in the reaction with the probe-based assay. This effect might be related to differences in the sensitivity to inhibitors of the Transcriptase reverse and/or Taq Polymerase included on each master mixes . 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 June 3, 2020. . https: //doi.org/10.1101 /2020 .05.30.20117721 doi: medRxiv preprint (Levesque-Sergerie et al., 2007 . The presence of inhibitors of the RT-qPCR reaction is a recurrent problem in clinical samples and the effect on the result of the reaction will depend on the enzymes (i.e. Transcriptase Reverse (M-MLV) and DNA Polymerase (Taq Polymerase)) used (Schrader et al., 2012) . Since the dye-based assay was not able to perform adequately using this approach, we do not recommend the use of this protocol in pooled pre-RNA extraction samples.

For massive surveillance use, we are confident to suggest the use of the probe-based assay in a pre-RNA extraction pooling testing up to 10 samples. For a system of pooling post-RNA extraction, we suggest first the use of probe-based assay followed by dye-based assays, and both can be used in pools up to 20 samples. Implementing a pool system for population sampling results in an important savings of laboratory resources and time, which are two key factors during an epidemic outbreak, such as for COVID-19, that may limit the surveillance approach selected for viral detection testing in the population. Further, the individual testing use, the dye-based performed similarly to the probe-based assays and could be used in case of a shortage of probe resources.

COVID-19 is a highly contagious disease causing devastating effects worldwide. Effective interventions are mandatory to control the transmission and spread of the virus. Using the pooling approaches evaluated here, we are confident that it can be used as a valid alternative assay for the detection of SARS-CoV-2 in human samples. The low cost and the potential use in pre-RNA extraction pool systems, place of this assays as a valuable resource for scalable sampling to larger populations such as surveillance targeting asymptomatic and presymptomatic individuals where massive testing is essential for the rapid identification of potential spreaders. Table 1 . Primers for nucleocapside amplification of SARS-CoV-2 by a dye-based assay used in this study.

A https://www.sigmaaldrich.com/covid-19.html Table 2 . Cycles threshold (CT) obtained from six 10-fold dilutions using the probe-based and dye-based real-time RT-PCR assays. ND= Not detected. Table 3 . Intra-assay repeatability and inter-assay reproducibility for the dye-based real-time RT-PCR assays. . 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 June 3, 2020 . . https://doi.org/10.1101 /2020 Table 4 . Cycles threshold (CT) obtained from four different pool post-RNA extraction settings using a probe-based (Probe) and dye-based (Dye) RT-qPCR assay. The original (positive) sample is denoted as 1. . 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 June 3, 2020. . https://doi.org/10.1101/2020.05.30.20117721 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 June 3, 2020. . https://doi.org/10.1101/2020.05.30.20117721 doi: medRxiv preprint Table 7 . Cycles threshold (CT) obtained from pools of 5 and 10 pre-RNA extraction samples using 5 and 10 uL of RNA template in a dye-based RT-qPCR assay. The original sample is denoted as individual CT. ND= Pools with not detectable amplification signal or melting at nonexpected temperature. . 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 June 3, 2020. 

Comprehensive detection and identification of human coronaviruses, including the SARSassociated coronavirus, with a single RT-PCR assay

The proximal origin of SARS-CoV-2

Presumed Asymptomatic Carrier Transmission of COVID-19

Pooled RNA extraction and PCR assay for efficient SARS-CoV-2 detection

CDC, 2020. Coronavirus Disease

Ad hoc laboratory-based surveillance of SARS-CoV-2 by real-time RT-PCR using minipools of RNA prepared from routine respiratory samples

Covid-19 mass testing facilities could end the epidemic rapidly

Universal weekly testing as the UK COVID-19 lockdown exit strategy

Ending the epidemic: Critical role of primary HIV infection

Identification of a novel coronavirus causing severe pneumonia in human: a descriptive study

Efficient high throughput SARS-CoV-2 testing to detect asymptomatic carriers

PCR inhibitors -occurrence, properties and removal

Comparison of SYBR Green and TaqMan methods in quantitative real-time polymerase chain reaction analysis of four adenosine receptor subtypes

Structure, Function, and Antigenicity of the SARS-CoV-2 Spike Glycoprotein

Limits of Detection of Six Approved RT-PCR Kits for the Novel SARS-coronavirus-2

Coronavirus disease 2019 (COVID-19) Situation Report -90. World health Organization (WHO)

Q&A: Similarities and differences -COVID-19 and influenza. World Health Organization

Development of a Laboratory-safe and Low-cost Detection Protocol for SARS-CoV-2 of the Coronavirus Disease 2019 (COVID-19)

Characteristics of and Important Lessons from the Coronavirus Disease 2019 (COVID-19) Outbreak in China: Summary of a Report of 72314 Cases From the Chinese Center for Disease Control and Prevention

SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients

We thanks Gobierno Regional de Los Rios and Universidad Austral de Chile for the financial support. Also, we thanks Dr. Marcela Perez (Hospital Lanco), Mr. Luis Leyton and Dr. Maritza Navarrete (Hospital Base de Valdivia), Dr. Omar Ulloa (Ejército de Chile), and Dr. Claudio Henríquez (Universidad Austral de Chile) for logistic and technical support.