key: cord-0939518-9c26we2m
authors: Anantharajah, Ahalieyah; Helaers, Raphael; Defour, Jean-Philippe; Olive, Nathalie; Kabera, Florence; Croonen, Luc; Deldime, Francoise; Vaerman, Jean-Luc; Barbée, Cindy; Bodéus, Monique; Scohy, Anais; Verroken, Alexia; Rodriguez-Villalobos, Hector; Kabamba-Mukadi, Benoit
title: How to choose the right real-time RT-PCR primer sets for the SARS-CoV-2 genome detection?
date: 2021-05-24
journal: J Virol Methods
DOI: 10.1016/j.jviromet.2021.114197
sha: e98ae09dcf3a13c938d480df70c77ca6292fe95b
doc_id: 939518
cord_uid: 9c26we2m

OBJECTIVES: The SARS-CoV-2 pandemic has created an unprecedented need for rapid large-scale diagnostic testing to prompt clinical and public health interventions. Currently, several quantitative reverse-transcription polymerase chain reaction (RT-qPCR) assays recommended by the World Health Organization are being used by clinical and public health laboratories and typically target regions of the RNA-dependent RNA polymerase (RdRp), envelope (E) and nucleocapsid (N) coding region. However, it is currently unclear if results from different tests are comparable. This study aimed to clarify the clinical performances of the primers/probe sets designed by US CDC and Charité/Berlin to help clinical laboratories in assay selection for SARS-CoV-2 routine detection METHODS: We compared the clinical performances of the recommended primers/probe sets using one hundred nasopharyngeal swab specimens from patients who were clinically diagnosed with COVID‐19. An additional 30 “pre-intervention screening” samples from patients who were not suspected of COVID-19 were also included in this study. We also performed sequence alignment between 31064 European SARS-CoV-2 and variants of concern genomes and the recommended primer/probe sets. RESULTS: The present study demonstrates substantial differences in SARS-CoV-2 RNA detection sensitivity among the primer/probe sets recommended by the World Health Organization especially for low-level viral loads. The alignment of thousands of SARS-CoV-2 sequences reveals that the genetic diversity remains relatively low at the primer/probe binding sites. However, multiple nucleotide mismatches might contribute to false negatives. CONCLUSION: An understanding of the limitations depending on the targeted genes and primer/probe sets may influence the selection of molecular detection assays by clinical laboratories.

Efforts to control SARS-CoV-2, the novel coronavirus causing COVID-19 pandemic, depend on accurate and rapid diagnostic testing. The reverse transcription real-time polymerase chain reaction (RT-qPCR) assay has become the gold standard for the diagnosis of SARS-CoV-2 infection. The European Commission advised to follow one of the World Health Organization protocols of RT-qPCR assays, published as early as January 2020 [1, 2] . Among them, the United States Center for Disease Control (US CDC) recommended two nucleocapsid gene targets (N1 and N2) [3] while the German Consiliary Laboratory for Coronaviruses hosted at the Charité in Berlin (Charité/Berlin) recommended first line screening with the envelope (E) gene assay followed by a confirmatory assay using the RNA-dependent RNA polymerase (RdRp) gene, even before the first COVID-19 cases appeared in Europe [4, 5] . At the time of data submission 437 molecular assays are commercially available or in development for the diagnosis of COVID-19 (https://www.finddx.org/) and most of them use these recommended gene targets alone or in combination. However, there has been no indication that any one of these sequence regions offer an advantage for clinical diagnostic testing, especially as the number of samples from patients with confirmed COVID-19 has been relatively small in the preliminary evaluations. Recent studies reported that the RT-qPCR assays have limited sensitivity, while chest computed tomography (CT) may reveal pulmonary abnormalities J o u r n a l P r e -p r o o f consistent with COVID-19, including ground-glass opacities, multifocal patchy consolidation, and/or interstitial changes with a peripheral distribution, even in patients with negative RT-qPCR results [6] [7] [8] . The genomic evolution of SARS-CoV-2 and specifically, the emergence of three new viral variants, at the end of 2020, associated with extensive transmission started to raise concerns [9] [10] [11] [12] . These variants of concern were first reported in the UK (B.1.1.7), South Africa (B.1.351) and Brazil (P.1). Their designation as VOCs was determined by an increase of local cases and by the high number of amino acid substitutions harboured by these lineages, which can lead to increased infectivity and potentially decreased vaccine effectiveness.

Genetic variations in the viral genome at primer/probe binding regions could result in potential mismatches and false-negative results. RT-qPCR assays with higher sensitivity targeting conserved regions might help to reduce the false-negative rate

In the context of lift confinement restrictions where large scale COVID-19 testing should be needed, this study aimed to clarify the clinical performances of the primer/probe sets designed by US CDC and Charité/Berlin to help guide assay selection by clinical laboratories for SARS-CoV-2 routine detection.

The amplified regions of SARS-CoV-2 by the recommended primers (E_Sarbeco_F1, E_Sarbeco_R2, E_Sarbeco_P1; RdRP_SARSr-F2, RdRP_SARSr-R1, RdRP_SARSr-P2; 2019-nCoV_N1-F, 2019-nCoV_N1-R, 2019-nCoV_N1-P; 2019-nCoV_N2-F, 2019-nCoV_N2-R, 2019-nCoV_N2-P) were aligned with SARS-Coronavirus, MERS-Coronavirus and seasonal human coronaviruses genome using MUSCLE [13] , and formatted using MSAViewer [14] .

Phylogenies have been inferred using MetaPIGA 3.1 [15] with "Human CoV 229E" selected as an outgroup. Resulting consensus trees have been formatted using iTOL.

In addition, 31064 SARS-CoV-2 sequences from 32 European countries and 4078 wellcharacterized SARS-CoV-2 sequences from 173 countries in the World have been downloaded from Global Initiative on Sharing All Influenza Data (GISAID) [16] , and aligned against N1, N2, RdRp and E primers/probe sequences using the Smith-Waterman algorithm [17] .

From April 1 to April 30, 2020, we retrospectively collected one hundred nasopharyngeal swab specimens from patients who were clinically diagnosed with COVID-19 according to the chest CT image and hospitalized in COVID-19 care units of the Cliniques universitaires Saint-Luc, in Brussels, Belgium. An additional 30 "pre-intervention screening" samples from patients who J o u r n a l P r e -p r o o f were not suspected of COVID-19 were also included in this study. The nasopharyngeal swabs were eluted in 1 mL of universal viral transport media (UTM, Copan, Italia). Nucleic acids were extracted from 650 µL of nasopharyngeal swab medium by the Abbott m2000sp following manufacturer's magnetic microparticle-based protocol (Abbott molecular, IL, USA). Samples were eluted in 60μL of elution buffer.

The reaction mix (20 μL) consisted of 4x TaqPath TM 

No significant homologies of SARS-CoV-2 sequences with MERS-CoV and seasonal human coronaviruses HKU1, OC43, NL63, 229E were observed suggesting a low risk of potential false positive RT-qPCR results with other circulating human coronaviruses. However, due to the relative paucity of positive control materials at the time these assays were developed, the primers and probe were designed such that they would also cross-react with the SARS-CoV [4, 18] . Indeed, probe and primers sequences of E gene assay showed high sequence homology with other related betacoronaviruses such as SARS coronavirus and bat SARS-like coronavirus genomes [4] . Although the forward and reverse primer sequences of RdRp and N2 assays showed also high sequence homology with SARS coronavirus, the combination of primers and probe would allow the specific detection of SARS-CoV-2. The assay targeting N1 gene was found to be specific to SARS-CoV-2.

To assess if these recommended primers and probes covered the circulating strains in Europe, 31064 SARS-CoV-2 sequences were downloaded from GISAID [16] , and compared to the primers and probe binding regions [17] (Table 1A) 

The RT-qPCR efficiencies of the recommended primer/probe sets, calculated using 10-fold serial dilution of RNA transcripts standard were above 90% which match the criteria for an efficient RT-qPCR assay [19] . The limit of detection was 5 copies per reaction for the N1 and N2 gene assays, 10 copies per reaction for the E gene assay and 25 copies per reaction for the RdRp-P2 assay ( Table 2) We collected nasopharyngeal samples from one hundred patients who presented Chest CT abnormalities consistent with COVID-19 (median age, 63.5 years; 51% female). The average length of stay in COVID-19 care units was 12 days with 15 deaths as of April 30 th 2020. The nasopharyngeal specimens were obtained upon hospital admission corresponding to an average of 5.6 days (range: 1-16) after symptoms onset.

Mean Ct values of the sample cohort detected by all RT-qPCR assays were significantly lower in N1 and N2 gene assays than in RdRp-P2 and E gene assay (One way ANOVA, Tukey posttest p<0.001). Compared to CT-Scan, N1 and N2 primer/probe sets showed the highest positive rate (73 and 74% respectively) followed by E primer/probe set (58%) and then RdRp primer/probe sets (44%) ( Table 2 ). The use of Genesig commercial kit with optimized target J o u r n a l P r e -p r o o f region and primer/probe sequences in the RdRp gene exhibited a slightly increased sensitivity (53%) compare to RdRp-P2 assay recommended by Corman et al [4] .

Interestingly, the combination of N1 and N2 assays allowed an increase in the sensitivity (84%) compared to N1 or N2 alone, including SARS-CoV-2 RNA detection in 5 additional specimens (viral load range: 519 -1007 copies/mL) that were tested negative by the others assays ( Figure   2 ). Among these patients, we could exclude 3 false positive results as patients had positive SARS-CoV-2 IgG (MAGLUMI assay, Snibe) 7-12 days after molecular testing. Serological control could not be performed for the 2 other patients as they died within a short time spanning.

The performances of the RT-qPCR assays were highly dependent on the viral load. In our study, positive SARS-CoV-2 clinical samples exhibited median Ct values > 30 corresponding to low viral load which made the detection challenging. Indeed, both RdRp and E assays reliably detected specimens with Ct values < 30, but did not detect 40-60% of specimens with Ct values ≥ 30 (Table 2, Figure 2 , Figure S1 ).

The survey published on February 11, 2020 reported [5] that the E-, RdRp-and N-gene assays had rapidly been implemented by the European laboratories. Very few studies have been published to date on the relative performance characteristics of these assays recommended by the WHO [3, 4, 20] . One of the key factors determining detection sensitivity is how efficiently primers and probes bind target genes. Our findings highlight substantial differences in sensitivity for the primer/probe sets when comparing under the same conditions. Indeed, N1

and N2 assays stand out in comparison with the E and RdRp assays for the detection of lowlevel viral loads. Furthermore, positive E and negative RdRp results were obtained in 15 cases.

We may therefore question the need of confirmatory testing following an initial positive test according to the Charité/Berlin protocol, resulting in turnaround time delay and increased workload. The cross-reactivity of the primers and probes with SARS-CoV should not cause any diagnostic ambiguity as SARS-CoV is no longer detected since the resolution of the epidemic in 2004 [21] . Based on our preliminary observations, multiplexing CDC N1 and N2 assays within a single PCR mixture could allow a reliable SARS-CoV-2 detection. For now, the low variability in the primer/probe binding regions of the SARS-CoV-2 sequences analysed, in particular of the B.1.1.7 VOC sequences can be considered reassuring. However, we observed notable mismatches in regions targeted by the primers/probe sets which might affect RT-qPCR assays performance depending on their location and the nature of the substitution [22, 23] .

At present, in the context of large-scale screening, RT-qPCR testing remains the standard for the diagnosis of COVID-19 despite the false-negative rate. Indeed, normal chest CT scan cannot exclude COVID-19, especially for patients with early symptoms [24] and conversely an abnormal CT scan is not specific for COVID-19 diagnosis [25] . We believe our results would encourage the laboratory staff to be aware of certain limitations depending on the targeted genes and to evaluate the clinical performances of COVID-19 molecular diagnostic tests across a wide range of viral concentrations before their implementation. Due to the extensive transmission of this virus, the genetic diversity in the SARS-CoV-2 genome and the constant emergence of variants of concern, we also encourage the detection of SARS-CoV-2 by targeting at least two distinct regions and oligonucleotide binding regions should be monitored continuously with the circulating virus strains.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper Table 1B : Primers alignments with variants of concern target sequences 4078 well-characterized SARS-CoV-2 sequences from 173 countries in the World have been downloaded from GISAID (Global Region-specific Auspice source files) and aligned against the sets of primers/probe. 356 different lineages (Pangolin nomenclature) and 12 or 9 different clades (Nextstrain and GISAID nomenclature respectively) were represented including the variants of concern (VOCs) in Europe (B. The viral load is expressed in log copies/mL and each clinical sample is represented by a circle. The white circles represent clinical samples detected by all RT-qPCR assays while colored circles represent samples not detected by the six assays. Bars represent the median and 95% Confidence Interval 

COVID-19 Technical Guidance : In House Molecular Assays 2020

Current performance of COVID-19 test methods and devices and proposed performance criteria 2020

First Case of 2019 Novel Coronavirus in the United States

Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR

Laboratory readiness and response for novel coronavirus (2019-nCoV) in expert laboratories in 30 EU/EEA countries

Use of Chest CT in Combination with Negative RT-PCR Assay for the 2019 Novel Coronavirus but High Clinical Suspicion

Correlation of Chest CT and RT-PCR Testing in Coronavirus Disease 2019 (COVID-19) in China: A Report of 1014 Cases

Chest CT for Typical 2019-nCoV Pneumonia: Relationship to Negative RT-PCR Testing

Emergence of genomic diversity and recurrent mutations in SARS-CoV-2

Whole genome and phylogenetic analysis of two SARS-CoV-2 strains isolated in Italy in

International expansion of a novel SARS-CoV-2 mutant

One year into the pandemic: Short-term evolution of SARS-CoV-2 and emergence of new lineages

The EMBL-EBI search and sequence analysis tools APIs in 2019

MSAViewer: interactive JavaScript visualization of multiple sequence alignments

MetaPIGA v2.0: maximum likelihood large phylogeny estimation using the metapopulation genetic algorithm and other stochastic heuristics

Global initiative on sharing all influenza data -from vision to reality

Identification of common molecular subsequences

Molecular Diagnosis of a Novel Coronavirus (2019-nCoV) Causing an Outbreak of Pneumonia

How good is a PCR efficiency estimate: Recommendations for precise and robust qPCR efficiency assessments

Rapid establishment of laboratory diagnostics for the novel coronavirus SARS-CoV-2 in Bavaria

SARS-CoV infection in a restaurant from palm civet. Emerg Infect Dis

The impact of primer and probetemplate mismatches on the sensitivity of pandemic influenza A/H1N1/2009 virus detection by real-time RT-PCR

The effect of primer-template mismatches on the detection and quantification of nucleic acids using the 5' nuclease assay

Patients with RT-PCR-confirmed COVID-19 and Normal Chest CT

Chest Computed Tomography for Detection of Coronavirus Disease 2019 (COVID-19): Don't Rush the Science