key: cord-283109-ka3n9pft authors: Arumugam, Arunkumar; Wong, Season title: The Potential Use of Unprocessed Sample for RT-qPCR Detection of COVID-19 without an RNA Extraction Step date: 2020-04-08 journal: bioRxiv DOI: 10.1101/2020.04.06.028811 sha: doc_id: 283109 cord_uid: ka3n9pft Quantitative reverse transcription polymerase chain reaction (RT-qPCR) assay is the gold standard recommended to test for acute SARS-CoV-2 infection.1–4 It has been used by the Centers for Disease Control and Prevention (CDC) and several other companies in their Emergency Use Authorization (EUA) assays. With many PCR-based molecular assays, an extraction step is routinely used as part of the protocol. This step can take up a significant amount of time and labor, especially if the extraction is performed manually. Long assay time, partly caused by slow sample preparation steps, has created a large backlog when testing patient samples suspected of COVID-19. Using flu and RSV clinical specimens, we have collected evidence that the RT-qPCR assay can be performed directly on patient sample material from a nasal swab immersed in virus transport medium (VTM) without an RNA extraction step. We have also used this approach to test for the direct detection of SARS-CoV-2 reference materials spiked in VTM. Our data, while preliminary, suggest that using a few microliters of these untreated samples still can lead to sensitive test results. If RNA extraction steps can be omitted without significantly affecting clinical sensitivity, the turn-around time of COVID-19 tests and the backlog we currently experience can be reduced drastically. Next, we will confirm our findings using patient samples. If RNA extraction steps can be omitted without significantly affecting clinical sensitivity, the turnaround time of COVID-19 tests and the backlog we currently experience can be reduced drastically. Next, we will confirm our findings using patient samples. The sample preparation step is generally time-consuming, regardless of whether it is done manually or automated. In addition, there is a current shortage of the recommended viral RNA extraction kits needed for the Centers for Disease Control and Prevention (CDC) RT-qPCR assay to diagnose SARS-CoV-2. During a study in developing a rapid protocol for influenza (Inf) and respiratory syncytial virus (RSV) diagnostics, we investigated the feasibility of omitting the sample preparation steps to expedite the test without significantly impacting the test's sensitivity. Using Inf and RSV clinical specimens, we successfully performed RT-qPCR reactions by simply adding a few microliters of the unprocessed sample in viral transport medium (VTM) directly into the RT-qPCR assay master mix. We then tested the approach using SARS-CoV-2 plasmid and SeraCare AccuPlex reference materials. The data presented below suggest that it is possible to skip the RNA extraction step in COVID-19 testing without a significant drop in assay sensitivity. We first tested the feasibility using a very small amount of sample in RT-qPCR reactions by using aerosol generating vials to spray the samples over the uncapped PCR tubes. The material in vials containing influenza A (InfA), influenza B (InfB), or RSV clinical specimens (swabs in VTM) were sprayed into PCR tubes containing primer and probe sets targeting InfA, InfB, RSV and RNaseP (RP) in master mix prior to capping the tubes and performing RT-qPCR. Fig. 1 shows that there were positive PCR signals in the respective tubes that did not have any non-specific amplification. Though the unprocessed, sprayed samples have higher Ct values (36.5, 36.2 and 31 for InfA, InfB, and RSV, respectively) than the corresponding extracted RNA template samples (30.4, 32.6, 29.6 for InfA, InfB, and RSV). It is important to note that we were able to detect low viral load samples (Ct >30) when less than 1 µL of sample entered the tubes as measured by an analytical balance. To test the extent of PCR inhibition exerted by flu specimens in VTM, we used clinical samples from Discovery Life Sciences, a Biospecimen repository. Between 0.1 µL to 6 µL of the unprocessed samples were directly spiked into the master mix with InfA primers in a 20 µL reaction mix (Fig. 2 ). An extracted template (in Promega Maxwell device) from the same sample was also amplified along with unprocessed samples. Adding more unprocessed samples (up to 6 µL) improved (reduced) the Ct values. The Ct difference between the extracted template (100 µL input and 100 µL eluate) and the unprocessed sample is minimal (29.3 vs. 30.3, respectively). Adding more than 6 µL of untreated samples has resulted in high Ct (data not shown), meaning the inhibitory effect outweighed the benefits of having more copies of the target in a reaction. We next tested whether the RNA from SARS-CoV-2 can be detected by directly spiking samples of the non-replicative recombinant virus particles (SeraCare AccuPlex SARS-CoV-2 reference material) in VTM to master mix without an extraction step. The SARS-CoV-2 virus particles were mixed with VTM to get a final concentration of 2,500 copies per mL. Different amounts (2, 4, 6 and 8 µL) of these mock clinical samples were spiked into the master mix containing CDC recommended SARS-CoV-2 RT-qPCR diagnostic panel primers N1, N2 o N3 in 20 µL PCR reactions, though we note that the N3 primers were recently removed by the CDC. Purified nucleic acid template (4 µL) isolated from a Promega Maxwell device was also amplified. The Ct difference between the purified template and the directly spiked clinical samples (i.e., 4 µL, 5 µL and 6 µL) is minimal. Then, 100 µL of AccuPlex SARS-CoV-2 reference material was processed in Promega Maxwell device (using an AS1520 cartridge) and eluted in 100 µL of elution buffer. The extracted template (4 µL) was added to master mix, and RT-qPCR amplification was carried out in a Bio-Rad (CFX-96) thermal cycler along with the reactions with unprocessed samples. As shown in Fig. 3 , the SARS-CoV-2 RNA from directly spiked samples was successfully detected by the RT-qPCR reaction without a nucleic acid extraction step (N1 target shown). Up to 8 µL of VTM and AccuPlex SARS-CoV-2 mix was used in the reaction, which did not exert major template (20 copies/reaction) is 35.5, which is 2 cycles lower than the result from 4 µL of untreated sample (10 copies) in the reaction (Fig. 3) . We speculate that the SeraCare matrix, containing Trisbuffered saline, glycerol, anti-microbial agents and human proteins, could be the reason for the above noticed PCR inhibition. We also tested VTM mixed SARS-CoV-2 plasmid (purchased from Integrated DNA Technologies, Inc.) using RT-qPCR. This plasmid is used as a positive control for the CDC's SARS-CoV-2 RT-qPCR assay. The positive control plasmid was mixed with VTM and 4 µL of this mix was used for the RT-qPCR reaction. The RT-qPCR results showed that the Ct values of control (without VTM) and VTM mixed reactions (all containing 400 copies/reaction) were very similar for all three (N1, N2 and N3) SARS-CoV-2 targets ( Table 3) . We also tested 40 copies of SARS-CoV-2 plasmid using RT-qPCR, which gave a Ct value of ~38 (data not shown). This means direct spiking of a higher concentration of SARS-CoV-2 genome copies in VTM did not have a major PCR inhibitory effect in the reactions when the target concentration is high ( Table 3) . Therefore, it is likely that specimens with higher viral load could also be detected by RT-qPCR without needing an RNA extraction step. SARS-CoV-2 plasmid in VTM did not show any PCR inhibition when compared to plasmid alone. The CDC positive control plasmid for SARS-CoV-2 (from IDT) was diluted in VTM or TE buffer (control) to get 100,000 copies per mL. 4 µL of SARS-CoV-2 plasmids in VTM and in TE buffer were added to a 20 µL PCR reaction mix targeting N1, N2 and N3 genes for the detection of SARS-CoV-2. The Ct values of both the control and plasmid in VTM are very similar, indicating that the sample preparation step can be omitted in high viral load samples. Our data using both high (Ct < 35) and low (Ct >35) target concentrations suggest that since RT-qPCR is highly sensitive, using raw samples or minimal sample preparation steps might not reduce the test sensitivity as most patients tend to have a higher viral load. 5 Also, the efficiency of extraction methods tends to drop significantly at very low target concentrations when processed through numerous washing steps. Therefore, nucleic acid loss during extraction steps may hurt the limit of detection. While further studies with patient specimens and a higher number of samples are needed to confirm our preliminary results, we report that the use of untreated samples can be a viable option during the COVID-19 pandemic. The AccuPlex SARS-CoV-2 reference material from SeraCare containing con-replicative viral particles (5,000 viral particles per mL) was mixed with equal volumes of VTM to get a final concentration of 2,500 viral particles per mL. Each microliter contained ~2.5 genomic material equivalents of SARS-CoV-2. SARS-CoV-2 positive control plasmid (CDC recommended) was obtained from Integrated DNA Technologies (200,000 copies/µL). This was diluted to 10,000 copies/µL in TE buffer and copies/mL. A control was prepared by diluting 10 µL of plasmid stock solution in 990 µL of TE buffer. Both the VTM and TE buffer working solutions (4 µL or 400 copies in each reaction) were added into a 20 µL PCR reaction. The Promega Maxwell extraction was performed using AS1520 cartridge (Promega Corporation, Madison, WI). 100 µL of the sample (AccuPlex SARS-CoV-2 or Influenza A/B or RSV) was added to the cartridge and eluted in 100 µL elution buffer provided in the kit. The authors declare no conflict of interest. Molecular Diagnosis of a Novel Coronavirus (2019-nCoV) Causing an Outbreak of Pneumonia A Novel Coronavirus from Patients with Pneumonia in China Comparison of different samples for 2019 novel coronavirus detection by nucleic acid amplification tests Detection of 2019 novel coronavirus (2019-nCoV) by real-time RT-PCR SARS-CoV-2 Viral Load in Upper Respiratory Specimens of Infected Patients Evaluation of Three Influenza A and B Real-Time Reverse Transcription-PCR Assays and a New 2009 H1N1 Assay for Detection of Influenza Viruses The Burden of Hospitalized Lower Respiratory Tract Infection due to Respiratory Syncytial Virus in Rural Thailand Division of Viral Diseases