Evaluation of alternative RNA extraction methods for detection of SARS-CoV-2 in nasopharyngeal samples using the recommended CDC primer-probe set

Background The efficiency of isolation and purification of the viral genome is a critical step to the accuracy and reliability of RT-qPCR to detect SARS-CoV-2. However, COVID-19 testing laboratories were overwhelmed by a surge in diagnostic demand that affected supply chains especially in low and middle-income facilities. Objectives Thus, this study compares the performance of alternative methods to extraction and purification of viral RNA in samples of patients diagnosed with COVID-19. Study design Nasopharyngeal swabs were submitted to three in-house protocols and three commercial methods; viral genome was detected using the primer-probe (N1 and N2) described by CDC and viral load of samples were determined. Results The in-house protocols resulted in detection of virus in 82.4 to 86.3% of samples and commercial methods in 94.1 to 98%. The disagreement results were observed in samples with low viral load or below the estimated limit of detection of RT-qPCR. Conclusion The simplified methods proposed might be less reliable for patients with low viral load and alternative commercial methods showed comparable performance.


Highlights
 Methods for viral RNA extraction with similar performance using CDC primer-probe set.
 Alternative in-house methods for SarsCov-2 detection under emergency situations.
 In-house and commercial methods with comparable SarsCov-2 detection by RT-qPCR.

BACKGROUND 1
Diagnosis of infected individuals is the cornerstone to track transmission and guide 2 strategies against the COVID-19 pandemic (Corman et al., 2020). Therefore, the 3 pandemic led to an unprecedented demand for diagnostic tests, overwhelming 4 laboratories especially in low and middle-income countries. 5 6 OBJECTIVES 7 Here, we aimed to investigate the performance of three in-house simplified and three 8 alternative commercial methods for viral RNA extraction and detection of SARS-CoV-9 2 in samples of patients with COVID-19. 10 11 STUDY DESIGN 12 Fifty-one nasopharyngeal swabs from individuals infected with SARS-CoV-2 were 13 selected according to cycle threshold (CT) in three groups of seventeen samples (CT 14 <20, between 20-30, and >30). For each extraction protocol 150 µL of PBS containing 15 the nasopharyngeal samples were used. For simplified protocols three microtubes were 16 used: first tube immediate incubated at 95°C for 10 minutes; second tube incubation 17 with 20 µL of proteinase K [20 mg/mL] (Promega, USA) at 56°C for 10 minutes, 18 followed by heating at 95°C for 10 minutes; thirty tube 150 µL of Chelex100 [10%] 19 (Sigma-Aldrich, USA) was added followed by vigorous vortexing, then incubation at 20 95°C for 10 minutes. After heating, all tubes were centrifuged and placed in an ice-bath 21 and then used for RT-qPCR assays. For purification using commercial kits three distinct  The Chelex 100 protocol resulted in increased CT for all assays. For RNAse P assay the 5 commercial kits showed lower CTs compared to in-house protocols. However, there were 6 no significant differences in the CT values or viral quantification (N1 assay and N2 assay) 7 among the different protocols ( Figure 1). 8 Comparing each sample result according to the RNA extraction/purification protocol, 9 eleven samples (21.6%) showed disagreement results in N1 or N2 assay. Those 10 inconclusive results and disagreement between different protocols were observed in 11 elutes with low yield. Since all elutes of clinical samples obtained from Sera-12 Xtracta®virus kit resulted in detectable N1 or N2 assays, showing the best performance 13 among protocols evaluated, we used the data of Sera-Xtracta®virus protocol to analyze 14 disagreement samples. The obtained CTs and quantifications of agreement and 15 disagreement samples showed that agreement among protocols is related to the viral load 16 in clinical samples (Figure 2).
The RT-qPCR performance is affected by the efficiency of the nucleic acids extractions 1 methods. Comparison of the available commercial kits in our laboratory showed not 2 significant differences in assays CT or viral load quantification. The automated 3 Maxwell® RSC 48 is recommended by CDC, but surprisingly showed lowest detection 4 rate, 94.1%, while the silica column-based method Reliaprep® viral TNA and beads-5 based method Sera-Xtracta® virus detected 98%. Indeed, detailed analysis showed that 6 concordance between commercial kits in samples with viral load above assays LOD is 7 100%. Accordingly, there was no significant difference between the manual or 8 automatized commercial extraction methods performance evaluated. assay. In our experiments, to inactivate viruses and reduce the risk of contamination for 16 simplified methods, the samples were heated for 10 minutes at 95°C prior to  That simplified protocol of heat-inactivation showed that viral RNA could be detected 18 from samples stored in PBS with agreement rates of 84% to 87.5% compared to 19 commercial extraction methods. To evaluate if the simplified protocol of heat-20 inactivation could be further improved, we added a previous step of proteinase K 21 treatment, which could enhance nucleic acids yield by inactivating RNAses present in 22 samples as well as other potential PCR inhibitors , and the incorporation 23 of Chelex-100, a chelating agent that binds to cellular components and stabilizes the RNA 24 during heat (Ulloa et al., 2020). However, unlike a previous study that observed increased detection when pre-treatment with proteinase K was employed (Chu et al, 2020), we were 1 unable to observe such improvement in simplified protocol of heat-inactivation. 2 The limitations of our study include, for example, the lack of a gold standard for SARS-3 CoV-2 detection, which lead to the calculation of agreement percentage, and the fact that 4 samples were collected and stored in PBS, and thus we could not observe the performance 5 of our simplified protocols in a distinct matrix of transport medium. Further, we could 6 not spike samples with known concentrations of synthetic genomic RNA, which could be 7 relevant to evaluate the efficiency of extractions/purifications methods. 8 Although we have used PBS for sample transport, we performed assays with reduced 9 amplicons sizes (71 bp and 67 bp for N1 and N2), samples were stored at -70°C, and RT-10 PCR was immediately performed after RNA extraction to avoid RNA degradation or kept 11 for short periods at -70°C and then analyzed by RT-PCR. The comparisons of commercial kits performance in this study support the 19 interchangeability of these methods and other factors such supply chain availability, cost, 20 and hands-on time should be evaluated to appoint the best RNA extraction/purification 21 method in each facility. The simplified extractions approach might be less robust and 22 affected by several conditions and should only be conducted under emergency use and 23 following proper validation. In conclusion, the data demonstrate that distinct extraction methods have comparable results in samples with viral load above the LOD or assay using 1 the CDC proposed primer-probe set. ☒ The authors declare that they have no known competing financial interests or 9 personal relationships that could have appeared to influence the work reported in this 10 paper. 11 12