key: cord-0785398-w4f7unne
authors: Xiu, Leshan; Binder, Raquel A.; Alarja, Natalie A.; Kochek, Kara; Coleman, Kristen K.; Son Than, Thé; Bailey, Emily S.; Bui, Vuong N.; Toh, Teck-Hock; Erdman, Dean D.; Gray, Gregory C.
title: A RT-PCR Assay for the Detection of Coronaviruses from Four Genera
date: 2020-04-30
journal: J Clin Virol
DOI: 10.1016/j.jcv.2020.104391
sha: 7bd1b1e6f4db991366f50e65f716fd47b2b1130b
doc_id: 785398
cord_uid: w4f7unne

Abstract Background During the past two decades, three novel coronaviruses (CoVs) have emerged to cause international human epidemics with severe morbidity. CoVs have also emerged to cause severe epidemics in animals. A better understanding of the natural hosts and genetic diversity of CoVs are needed to help mitigate these threats. Objective To design and evaluate a molecular diagnostic tool for detection and identification of all currently recognized and potentially future emergent CoVs from the Orthocoronavirinae subfamily. Study design and Results We designed a semi-nested, reverse transcription RT-PCR assay based upon 38 published genome sequences of human and animal CoVs. We evaluated this assay with 14 human and animal CoVs and 11 other non-CoV respiratory viruses. Through sequencing the assay's target amplicon, the assay correctly identified each of the CoVs; no cross-reactivity with the 11 common respiratory viruses was observed. The limits of detection ranged from 4 to 4 × 102 copies/reaction, depending on the CoV species tested. To assess the assay's clinical performance, we tested a large panel of previously studied specimens: 192 human respiratory specimens from pneumonia patients, 5 clinical specimens from COVID-19 patients, 81 poultry oral secretion specimens, 109 pig slurry specimens, and 31 aerosol samples from a live bird market. The amplicons of all RT-PCR-positive samples were confirmed by Sanger sequencing. Our assay performed well with all tested specimens across all sample types. Conclusions This assay can be used for detection and identification of all previously recognized CoVs, including SARS-CoV-2, and potentially any emergent CoVs in the Orthocoronavirinae subfamily.

• The MERS, SARS and COVID-19 epidemics have demonstrated that coronavirus (CoVs) have an affinity for interspecies transmission and causing epidemics.

• We developed a diagnostic to detect all CoVs from the four main genera.

• This assay can detect and identify all previously recognized CoVs and any future related CoVs that may emerge.

• The assay was highly specific and sensitive in detecting CoVs, and performed well on different sample types.

• Sustained surveillance for the emergence of novel CoVs is greatly needed at the human-animal interface. This diagnostic may help ABSTRACT Background: During the past two decades, three novel coronaviruses (CoVs) have emerged to cause international human epidemics with severe morbidity. CoVs have also emerged to cause severe epidemics in animals. A better understanding of the natural hosts and genetic diversity of CoVs are needed to help mitigate these threats. Objective: To design and evaluate a molecular diagnostic tool for detection and identification of all currently recognized and potentially future emergent CoVs from the Orthocoronavirinae subfamily. Study design and Results: We designed a semi-nested, reverse transcription RT-PCR assay based upon 38 published genome sequences of human and animal CoVs. We evaluated this assay with 14 human and animal CoVs and 11 other non-CoV respiratory viruses. Through sequencing the assay's target amplicon, the assay correctly identified each of the CoVs; no cross-reactivity with the 11 common respiratory viruses was observed. The limits of detection ranged from 4 to 4×102 copies/reaction, depending on the CoV species tested. To assess the assay's clinical performance, we tested a large panel of previously studied specimens: 192 human respiratory specimens from pneumonia patients, 5 clinical specimens from COVID-19 patients, 81 poultry oral secretion specimens, 109 pig slurry specimens, and 31 aerosol samples from a live bird market. The amplicons of all RT-PCR-positive samples were confirmed by Sanger sequencing. Our assay performed well with all tested specimens across all sample types. Conclusions: This assay can be used for detection and identification of all previously recognized CoVs, including SARS-CoV-2, and potentially any emergent CoVs in the Orthocoronavirinae subfamily.

Keywords: Coronavirus, emerging, infectious diseases, SARS-CoV-2, COVID-19

The family Coronaviridae consists of the largest single-stranded RNA viruses that infect humans, other mammals, and birds [1, 2] . The Coronaviridae is comprised of two subfamilies, Letovirinae, and Orthocoronavirinae. The Letovirinae subfamily includes only one previously recognized virus (partial genome) which was detected in [4] . Notably, SARS-CoV, MERS-CoV and SARS-CoV-2 infect humans, while SADS-CoV causes disease only in pigs [5, 6] .

While the other zoonotic CoV outbreaks have been largely controlled, attempts at containing SARS-CoV-2 have not been successful. Most experts would agree that this has been due, in part, to the sparse availability of diagnostic assays. Hence, currently, J o u r n a l P r e -p r o o f there is a massive, worldwide effort to provide diagnostic tests for SARS-CoV-2 infections. According to the Foundation for Innovative New Diagnostics website [7], as of April 9, 2020 there were more than 470 molecular and immunological assays in the development pipeline. However, few diagnostics are available or being developed to detect other animal CoVs that may jump species to man. Hence, in this study we sought to design and test a new molecular diagnostic to detect most all CoVs, including those that have been previously identified and those that will be discovered in the future.

In this work we sought to develop and test a semi-nested, RT-PCR assay that targets the conserved RNA-dependent RNA polymerase (RdRp) genome region common among all members of the Orthocoronavirinae, including SARS-CoV-2.

In designing the assay, we selected genome sequences of 38 representative viruses from α-, β-, γ-, and δ-CoV genera that were available in the National Center for Biotechnology Information (NCBI) GenBank database.

Since the RdRP is the most conserved region among the Orthocoronavirinae, this region was chosen for designing RT-PCR primers to all members of this subfamily. As a first step, RdRp sequences were extracted from NCBI and multiple alignments were J o u r n a l P r e -p r o o f prepared using the Clustal W program with default parameters implemented in MEGA 7 (http://www.megasoftware.net). Phylogenetic trees of the RdRp genes were then constructed using the neighbor-joining method with 1000 bootstrap replicates also using MEGA 7 (Fig 1) . The GenBank accession numbers of these viruses are recorded in Table S1 .

The 38 CoV RdRp gene sequences from GenBank were next aligned using Geneious software (Biomatters Ltd, Auckland, New Zealand). Highly conserved regions were visually identified and three degenerate primers were then designed using Geneious software to be broadly reactive across all Orthocoronavirinae. All primers were synthesized by Integrated DNA Technologies (IDT, Coralville, IA, USA). For first round RT-PCR, the external primers (pan-CoV_outF and pan-CoV_R) were used to amplify a 670-673 bp product; for the second round semi-nested PCR, an internal primer (pan-CoV_inF) was used with pan-CoV_R for amplification of an internal 599-602 bp product. The primer sequences used for each amplification are shown in Table S2 . Virginia, USA) model BC 251 two-stage bioaerosol sampler from a large poultry market in Vietnam [9] . Finally, we compared the pan-CoV assay's performance to the US Center for Disease Control and Prevention's 2019-nCoV RT-PCR assay [11] against a set of clinical specimens from COVID-19 patients.

The protocol consisted of a semi-nested RT-PCR strategy with two amplification steps. A first RT-PCR amplification round was conducted using SuperScript™ III 

Additional study methods such as RNA extraction methods, and construction of the recombinant plasmids used in testing are described in the supplemental materials. 

Through alignment of 38 RdRp gene sequences from human and animal CoVs, the region containing the most conserved sequences was identified (Fig 2) . Two degenerate primer pairs were constructed (Table S2) 

To validate the specificity of the pan-CoV assay, a wide range of different CoVs were prepared for templates. The assay specificity was further determined by attempting to amplify nucleic acid from non-specific pathogens like common respiratory viruses. Our assay yielded an amplicon of the expected size (599-602 bp) from the well-characterized control and reference CoV strains (Fig 3) . The results showed that the pan-CoV assay could accurately identify different CoVs without cross-reactivity with other non-target pathogens, demonstrating a high specificity for the primers used in this study (Fig 3) . Furthermore, Sanger sequencing confirmed the amplification of expected sequences from the tested samples (Table S3) . Some other

CoVs (except for the tested CoVs) were not available for testing, but in silico alignment of the targeted region in the RdRp gene of these species demonstrated a J o u r n a l P r e -p r o o f sufficient primer match to predict successful amplification.

To determine the limits of detection (LOD) of this assay, 10-fold dilution series of templates (plasmids and RNA) were tested, each performed in triplicate. The LOD was defined as the lowest concentration that was detected in all triplicate reactions.

Under general semi-nested RT-PCR conditions, our results revealed that the assay was able to detect bands in gels at dilutions ranging from 4 to 4×102 copies/reaction for corresponding plasmids (SARS-CoV with 4 and MERS-CoV with 4×102 copies/reaction). In addition, RNA extracted from SARS-CoV-2 (USA-WA1 isolate) infected cells with known concentration (with a pre-determined titer of 4.8×107 genome copies/ml, data not shown) was used to determine the LOD of this assay.

These data revealed that the pan-CoV assay could detect SARS-CoV-2 RNA at 4×101 copies per reaction.

Intra-and inter-assay variability for the assay were determined at high, medium and low copy number (106, 104 and 102) using the plasmids prepared above, which were serially diluted 10-fold to a final concentration between 1 and 108 copies/ L. For intra-assay variability, all samples were run in triplicate. For inter-assay variability, the assays were repeated three times individually on different days. The plasmids, with different dilutions, were successfully detected under different conditions and showed similar results among the assays. It is worth mentioning that this assay was also successfully implemented in the hands of a research team in Singapore (Fig. S1 in the supplemental materials).

To assess the performance of our semi-nested RT-PCR assay with human clinical specimens, a total of 192 swab samples from patients with pneumonia were tested. Of these, nine (4.7%) tested positive with the expected amplicon size (Table S4) To determine whether the assay could be used to identify naturally occurring CoVs from the environment, we blindly screened 31 bioaerosol samples collected from a poultry market in Hanoi, Vietnam. Among the 31 bioaerosol samples tested, three (9.7%) were positive for duck CoV, two (6.5%) were positive for infectious bronchitis virus and one (3.2%) was positive for avian CoV (Table S4) . Negative control samples were negative in all tests, ruling out amplicon contamination as a source of the positive results.

Finally, the pan-CoV assay had 100% agreement with the US CDC's rRT-PCR assay for SARS-CoV-2 on two nasopharyngeal swabs, two saliva specimens, and one rectal swab from COVID-19 patients (Table S5) .

J o u r n a l P r e -p r o o f

Detection of CoVs has most often been made by electron microscopy, cell culture, serological assays, or molecular methods. Microscopy requires expensive equipment and technical expertise. Although cell culture has been considered the gold standard for CoV detection with high specificity [12] , its time requirements and low sensitivity limit its clinical application. Serologic assessments are fraught with low sensitivity and occasional cross-reactivity [13] . While fast and sensitive [14] , molecular assays target previously recognized CoVs and may miss the detection of a novel CoV [15] .

Our pan-CoV assay overcomes these problems. It employs a semi-nested RT-PCR approach using two pairs of degenerate primers that universally amplify CoVs within four genera. We recognize that semi-nested RT-PCR methods are more complex and offer additional risk of assay contamination. However, we argue that effective management of the laboratory environment can reduce such risk.

As the COVID-19 pandemic is currently surging, pathogen discovery of novel CoVs now seems even more important than before. All HCoVs come from two main genera (α-CoV and β-CoV) and are thought to have come from animals, specifically bats [16] [17] [18] [19] . It seems clear that large-scale surveillance is needed to fully understand the role played by γ-CoVs and δ-CoVs in the emergence of human CoVs. Hence, this assay might be especially useful in detecting CoVs that are potentially pathogenic to humans from all four genera. Interestingly, in our study, viruses that are genetically similar to canine-CoV were detected in four human pneumonia samples. We are in the J o u r n a l P r e -p r o o f process of further investigating these findings with attempts at culture and next generation sequencing.

With the advent of next generation sequencing (NGS) technology [20] , numerous CoV strains of four genera have been identified via the increased surveillance of wild animal species. NGS has helped elucidate the evolutionary history of CoVs using a large amount of data generated by unbiased sequencing. However, some disadvantages limit the number of samples that can be sequenced. Although it becomes less expensive each consecutive year, the cost is still very high and the procedure includes time-consuming complicated library preparations and data processing. Hence, we posit that combined screening strategies are the most costeffective way to identify new CoVs. We recommend employing novel coronavirus surveillance among people whose occupations or recreation cause them to have prolonged or frequent exposure to animals. We recommend screening them periodically, when healthy, for novel CoV nasal carriage and whenever they develop an acute respiratory illness. This surveillance might first include employing our pan-CoV assay and other pan-viral assays. When a novel virus is detected, the surveillance team might look for that same virus among the workers' animals. Having confirmed a novel virus in both an animal worker and his or her animals, further investigations might be employed through cell culture, full genome sequencing, and if indicated, seroepidemiology of both the worker and the animals.

Novel CoVs also have caused tremendous morbidity among livestock. Currently, six CoVs have caused significant morbidity in pigs [5] . These include four α-CoVs Among them, SADS-CoV may have originated from bat CoVs and PDCoV from a sparrow CoV [5] , indicating that pigs are the likely complex reservoir for containing these emerging different CoVs. Given its ability to detect CoVs from four CoV genera, our pan-CoV assay can be used to detect novel CoVs in animal herds.

It should be noted that δ-CoVs in particular have an affinity to jump species. A recent study demonstrated that poultry are susceptible to infection with PDCoV [21] and that PDCoV also infects human cell lines [22] . So, we should be particularly concerned when a novel CoV is detected at the human-animal interface.

We have designed and evaluated a pan-CoV assay that can detect all known human CoVs from four genera. This assay may be used as a diagnostic tool to identify all previously recognized CoVs, including SARS-CoV-2 and related viruses, as well as any future emergent CoVs from the Orthocoronavirinae subfamily.

LX designed and tested the pan-CoV assay at Duke University, and drafted the manuscript. GCG conceived the study, guided the work, and helped draft the 

Middle East respiratory syndrome coronavirus: another zoonotic betacoronavirus causing SARS-like disease

Molecular Evolution of Human Coronavirus Genomes

Three Emerging Coronaviruses in Two Decades

Emerging and re-emerging coronaviruses in pigs

Fatal swine acute diarrhoea syndrome caused by an HKU2-related coronavirus of bat origin

High Prevalence of Viral Infections Among Hospitalized Pneumonia Patients in Equatorial Sarawak

Bioaerosol Sampling to Detect Avian Influenza Virus in Hanoi's Largest Live Poultry Market

A Feasibility Study of Conducting Surveillance for Swine Pathogens in Swine Slurry in North Carolina Swine Farms. Under journal review

CDC 2019-Novel Coronavirus (2019-nCoV) Real-Time RT-PCR Diagnostic Panel

Cell Culture, Technology: Enhancing the Culture of Diagnosing Human Diseases

Serological assays for emerging coronaviruses: challenges and pitfalls

Current and future molecular diagnostics for ocular infectious diseases

Establishment and Application of a Universal Coronavirus Screening Method Using MALDI-TOF Mass Spectrometry

Hosts and Sources of Endemic Human Coronaviruses

Evidence for an Ancestral Association of Human Coronavirus 229E with Bats

Surveillance of Bat Coronaviruses in Kenya Identifies Relatives of Human Coronaviruses NL63 and 229E and Their Recombination History

Evidence supporting a zoonotic origin of human coronavirus strain NL63

Bats and Coronaviruses

Porcine Deltacoronavirus Infection and Transmission in Poultry, United States

Broad receptor engagement of an emerging global coronavirus may potentiate its diverse cross-species transmissibility

The authors have declared that no competing financial interests exist.

We give special thanks to Dr. Gregory Sempowski, Director of the Duke Global Health Research Building, a.k.a. Regional Biocontainment Laboratory