One-tube SARS-CoV-2 detection platform based on RT-RPA and CRISPR/Cas12a

Background COVID-19 has spread rapidly around the world, affecting a large percentage of the population. When lifting certain mandatory measures for an economic restart, robust surveillance must be established and implemented, with nucleic acid detection for SARS-CoV-2 as an essential component. Methods We tried to develop a one-tube detection platform based on RT-RPA (Reverse Transcription and Recombinase Polymerase Isothermal Amplification) and DNA Endonuclease-Targeted CRISPR Trans Reporter (DETECTR) technology, termed OR-DETECTR, to detect SARS-CoV-2. We designed RT-RPA primers of the RdRp and N genes following the SARS-CoV-2 gene sequence. We optimized reaction components so that the detection process could be carried out in one tube. Specificity was demonstrated by detecting nucleic acid samples from pseudoviruses from seven human coronaviruses and Influenza A (H1N1). Clinical samples were used to validate the platform and all results were compared to rRT-PCR. RNA standards and pseudoviruses diluted by different gradients were used to demonstrate the detection limit. Additionally, we have developed a lateral flow assay based on OR-DETECTR for detecting COVID-19. Results The OR-DETECTR detection process can be completed in one tube, which takes approximately 50 min. This method can specifically detect SARS-CoV-2 from seven human coronaviruses and Influenza A (H1N1), with a low detection limit of 2.5 copies/µl input (RNA standard) and 1 copy/µl input (pseudovirus). Results of six samples from SARS-CoV-2 patients, eight samples from patients with fever but no SARS-CoV-2 infection, and one mixed sample from 40 negative controls showed that OR-DETECTR is 100% consistent with rRT-PCR. The lateral flow assay based on OR-DETECTR can be used for the detection of COVID-19, and the detection limit is 2.5 copies/µl input. Conclusions The OR-DETECTR platform for the detection of COVID-19 is rapid, accurate, tube closed, easy-to-operate, and free of large instruments.


Background
The outbreak of the novel severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has attracted the attention of people all over the world. By August 1, 2020, 17,396,943 cases and 675,060 deaths worldwide had been recorded [1]. In China, for example, although the spread of the epidemic has potentially been contained, there are still sporadic cases of cluster infection or imported coronavirus disease . Coexistence with the novel coronavirus will become a new normality. To ensure orderly production, life, work and study, large-scale, rapid and accurate viral detection is essential [2].
For SARS-CoV-2, laboratories around the world have developed a variety of viral nucleic acid detection platforms, such as sequencing [3,4], digital PCR [5], and the widely used real-time reverse transcription PCR (rRT-PCR), as recommended by China CDC and the WHO [6,7]. Generally, the sensitivity of rRT-PCR assays ranges from 3.8 to 10 RNA copies per reaction, with high specificities. These technical platforms require expensive equipment and professional technicians. Additionally, the testing processes take more than 1.5 h [8].
In vitro nucleic acid detection technologies based on clustered regularly interspaced short palindromic repeats (CRISPR)/Cas targeted recognition and promiscuous cleavage activities have been widely reported [9][10][11][12]. At present, the method of combining loop-mediated isothermal amplification (LAMP) with CRISPR/Cas12 [13] or RPA with CRISPR/Cas13 [14] has been used to detect SARS-CoV-2. Relatively complex primers must be designed in LAMP amplification, and more components added to the CRISPR/Cas13a testing mix. Since the reactions of isothermal amplification and CRISPR/Cas interfere with each other, the isothermal amplification step is performed first, after which a small amount of amplification product is transferred to the CRISPR/Cas detection system. That might easily produce aerosol contamination in the lab. These rapid platforms have low cost and equipment requirements. The highest sensitivity of these assays ranges from 1 to 4.5 RNA copies per μl [8].
In this study, OR-DETECTR platform was used to detect SARS-CoV-2. This platform might shorten detection time, reduce equipment cost and not produce aerosol contamination.

Human clinical sample collection and preparation
Pharyngeal swab samples were collected from 14 fever patients in Shenzhen Second Peoples' Hospital by the Clinic Diagnosis Laboratory and the nucleic acids of each sample were extracted with the pre-packaged nucleic acid extraction kit (Da'An Gene., Ltd.), according to the manufacturer's instructions, resulting in 55 μl extracts for each sample.
An additional 40 pharyngeal swab samples from 40 fever patients in Shenzhen Second Peoples' Hospital were also collected and extracted. rRT-PCR assay employing 5 μl of extract from each sample, were all negative. The remaining 50 μl extract of each sample was combined and 5 μl of the mixture was once again analyzed by rRT-PCR and confirmed to be SARS-CoV-2 negative. The RNA mix was named N50B.
Patients with fever underwent nucleic acid testing for two consecutive days. Patients who tested positive for two consecutive nucleic acid tests were transferred to designated hospitals for treatment. We continued to test patients who had negative tests and no ground glass shadow on chest X-ray to find the cause and determine treatment. The samples used in this study were nucleic acids extracted from patients' pharyngeal swabs on day one. These patients underwent two rRT-PCR tests and chest X-rays and were evaluated as negative.

OR-DETECTR assays
The OR-DETECTR platform combines RPA and CRISPR/Cas12a detection. One RPA pellet per tube was resuspended with 29.5 μl rehydration buffer in the TwistAmp basic RPA kit. RT-RPA reactions containing 5 µl of sample extract, 14.75 μl resuspended RPA solution, 1.2 μl of each primer (10 mΜ), 0.75 μl RNA reverse transcriptase (100 U/ml), 0.85 μl nuclease-free water and 1.25 μl MgAc (280 mM), with the total volume of 25 μl. The multiple RT-RPA reaction contains four primers, each of which is 0.6 μl (10 mΜ). After being gently mixed and centrifuged, the mixture was placed at the bottom of a centrifuge tube. The CRISPR/ Cas12a reaction mix consisted of 2 μl NEB Buffer 3.1 (10 ×), 2 μl LbCas12a (100 ng/μl), 2 μl crRNA (200 nM), 2 μl ssDNA reporter, and 12 μl nuclease-free water. This 20 μl CRISPR/Cas12a mix was added to the lid of the centrifuge tube before sealing. The tube was placed in a thermocycler or water bath at 42 °C and incubated for 30 min. The CRISPR/Cas12a mixture in the cap was combined with the RPA reaction product with a short spin. If ssDNA reporter was 5′-6-FAM-TTA TTA TT-BHQ1-3′, the centrifuge tube was then placed in the PE microplate reader at 42 °C, with the fluorescence signal collected every min for 20 min total. If ssDNA reporter was 5′-6-FAM-TTA TTA TT-Biotin-3′, the centrifuge tube was then placed in a thermocycler or water bath at 42 °C. The reaction product (3 μl) was diluted in 400 μl water, after which 80 μl was added to the sample hole of the test card (GenDx Suzhou).

rRT-PCR assays
Real-time RT-PCR reaction kit was supplied by Biogerm, which uses the primers recommended by the China CDC. Reactions were conducted in a 25 μl volume following the kit instructions. Reaction cycle parameters were set as reverse transcription at 50 °C for 10 min, denaturation at 95 °C for 5 min, followed by 40 cycles of amplification, 95 °C for 10 s and 55 °C for 40 s.

Statistical analysis
Statistical analyses were performed using GraphPad Prism. P < 0.001 was considered to indicate a statistically significant difference.

The workflow of the OR-DETECTR platform
Here, the OR-DETECTR was used for the detection of RNA extracted from pharyngeal swabs of COVID-19 patients. The detection process of the OR-DETECTR platform is shown in Fig. 1a. The RNA sample extracted from the pharyngeal swab is added to the RT-RPA mix at the bottom of the centrifuge tube while the DETECTR mix is located on the transparent tube lid. To reduce the use of the equipment, the reaction temperature for reverse transcription, amplification and detection is set at 42℃. Reverse transcription of viral RNA and highsensitivity amplification of specific sequences can be completed simultaneously in 30 min at 42 °C. RT-RPA and DETECTR components are physically separated during this amplification process. Following amplification, the DETECTR mix on the lid is mixed with amplification products by centrifugation. The amplicon can be recognized by specific CRISPR RNA (crRNA) segments, while the reaction activates the cis-and trans-nuclease activities of Cas12a. Cas12a nonspecifically cleaves the reporter sequence (quenched fluorescent ssDNA) and  [12]. We have shown that the one-tube detection platform based on RT-RPA and SHERLOCK, termed OR-SHERLOCK, also works. Preliminary comparison of OR-DETECTR and OR-SHERLOCK revealed that the detection effect was essentially the same, although components of the SHERLOCK mix are more complex (Additional file 1: Figure S1, S2).

The primers and crRNAs design of the OR-DETECTR platform
SARS-CoV-2 is an enveloped virus that contains a positive-sense single-stranded RNA genome of 29,903 nucleotides [15]. We designed primers targeting the RdRp (RNA-dependent RNA polymerase) and N (nucleoprotein) genes of SARS-CoV-2 (Fig. 1b). There is a partial overlap within the RdRp gene amplification region between OR-DETECTR assay and the World Health Organization (WHO) assay [16]. This RdRp gene of pan beta-CoV was designed as a reference assay [2,17]. The majority of our N gene amplification region overlaps the China-CDC assay (N gene assay) [6]. Including the required protospacer adjacent motif sites (PAM), we designed Cas12a crRNAs to detect RdRp and N gene amplicons. Additionally, we added a T7 promoter to the 5′ end of the forward primers, so that amplification products could also be detected by the OR-SHERLOCK platform (Additional file 1: Figure S2).  [4]. Partial RdRp and N gene sequences of all seven coronaviruses were aligned to the corresponding sequences of amplicons by ClustalW and GeneDoc (Fig. 1c, e). Influenza A is a group of highly contagious seasonal influenza viruses. It can be seen that the RdRp gene sequence and the N gene sequence of the SARS-CoV virus are most similar to SARS-COV-2. Using seven commercially coronaviruses and Influenza A (H1N1) at 200 copies/μl respectively, the OR-DETECTR can distinguish SARS-CoV-2 with no cross-reactivity from the other six coronavirus strains and H1N1 using the N gene target. It can also distinguish SARS-CoV-2 and SARS-CoV with expected cross-reactivity from the other five coronavirus strains and H1N1 with the RdRp gene target (Fig. 1d, f ). We included one positive control (PC: 80 copies per μl input) and one negative control (NC: no template) per trial and recorded the fluorescence signal value every min for 19 min. We defined the fluorescence signal value at 0 min as B and the fluorescence signal value at 19 min as F. In each test, the background-subtracted fluorescence was normalized by the negative control, and the fold change values (FC) were generated as FC = (F (PC) − B (PC))/(F (NC) − B (NC)).

OR-DETECTR platform test clinical samples
Next, we used the OR-DETECTR platform to test 14 RNA samples extracted from pharyngeal swabs, including six rRT-PCR-positive COVID-19 patients and eight rRT-PCR-negative patients with fever. N50B was a mixture of 40 negative samples, which was included as one sample and also tested via OR-DETECTR. Cycle Threshold (CT) values of N and Orf1ab genes were assessed using the SARS-CoV-2 test kit from BioGerm (Additional file 1: Table S1). We found that 4001 N was a positive sample (FC-RdRp is 1.9 and FC-N is 4.6) by OR-DETECTR, and CT values of 4001 N were 38.9 (Orf1ab) and 36.8 (N gene), meaning this clinical sample had very low viral loads. Although RT-PCR can be used for quantitative analysis and OR-DETECTR cannot, the judgment of negative and positive results between the two platforms was consistent when comparing CT and FC values ( Fig. 2a, b, Additional file 1: Fig. S3, Table S1). The negative samples yielded a maximal FC value of 0.8 with the RdRp gene and 1.3 with the N gene, whereas positive samples had a minimal FC value of 1.9 on RdRp and 4.6 on the N gene. Cut-offs (1.6 on RdRp gene and 2.6 on N gene) were set at two times the maximal FC value of negative samples.

Limit of detection of the OR-DETECTR platform
We next evaluated the limit of detection (LoD) of the OR-DETECTR system with SARS-CoV-2. A commercial RNA standard containing RdRp transcripts (1.1 × 10 6 copies/μl) and N gene transcripts (8.38 × 10 5 copies/μl) was serially diluted in TE buffer, with seven replicates at each concentration, and an RNA negative control of buffer alone. The estimated LoD for the OR-DETECTR is 10 copies per µl input with RdRp, as FC values of all seven replicates were greater than 1.6 (Fig. 2c). The LoD with the N gene transcript target is 2.5 copies per µl input, as FC values of all seven replicates were greater than 2.6 (Fig. 2d). The pseudovirus is better for simulating a clinical sample, furthermore, to avoid the use of largescale equipment and reduce costs, we used the method of heating the lysate to extract the pseudo-viral nucleic acid which had been diluted in gradients. We then evaluated the sensitivity of the OR-DETECTR platform. The LoD for the OR-DETECTR is 20 copies per µl input with RdRp, while the LoD for the N gene transcript target is 1 copy per µl input. The TritonX-100 in the lysate did not affect the nucleic acid detection of the OR-DETECTR platform (Additional file 1: Figure S4). The N gene target has superior sensitivity and specificity to RdRp, meaning the RdRp gene was designated as a reference assay.

Lateral flow strip detection based on OR-DETECTR
To avoid using large equipment, the OR-DETECTR assay can be visualized by a test card with a lateral flow strip. Activated-Cas12a cleaves FAM-biotin reporter and generates a signal at the first detection line (test line), while uncleaved-reporters are captured at the second detection line (that is, the control line). The OR-DETECTR assay is considered positive if there are two lines, negative if there is one control line, and meaningless if there is no control line. There were two replicates per dilution, and both 2.5 copies per µl input were classified as positive, but only one of 1.25 copies per µl input was positive. The LoD of the lateral flow strip is 2.5 copies per µl input, which is the same as the fluorescence plate reader (Fig. 2e). Additionally, the visual signal was achieved within 4 min by test card. Based on the lateral flow strip, we tried a multiple RT-RPA reaction. The RdRp gene and N gene were amplified in a RT-RPA mixture, after which the amplified product was detected in two CRISPR mixtures containing the crRNA specifically targeting the RdRp or N gene. The results were shown through two lateral flow strips. Results indicated that multiple RT-RPA detection was feasible and that when the sample concentration was 10 copies per µl input, both RdRp and N genes were positive (Additional file 1: Figure S5).

Discussion
Effective SARS-CoV-2 diagnosis is an important prerequisite for controlling the COVID-19 pandemic. The WHO proposed the ASSURED criteria (Affordable, Sensitive, Specific, User-friendly, Rapid and robust, Equipment-free, and Deliverable to end-users) to determine whether the diagnostic methods being used meet the needs of epidemic control [18]. However, few diagnostic methods meet all standards. For example, rRT-PCR testing is not fast enough, and many kits require qualified laboratories, expensive equipment and professional testing personnel.
We systematically compared the cost, turnaround time, sensitivity, specificity and so on of the current assay with existing COVID-19 diagnostic assays (Table 1).
Although rRT-PCR is currently the most commonly used diagnostic method for COVID-19, the LoD of those kits developed by different companies ranges from 100 copies/ml to 1000 copies/ml [8]. A real-time PCR system is a relatively complex, expensive instrument which not all hospitals could configure in their laboratory departments. This means that samples must be sent to the local CDC for testing, increasing testing time as well as various other risks. The detection method based on rRT-PCR is not flexible, since one instrument detects one 96-well plate at a time and the turnaround time is greater than 1.5 h. In contrast, the detection method based on isothermal amplification is flexible, greatly improves detection efficiency as the number of samples ranges from 1 to 384 [13], and the fluorescence detection time is less than 30 min at a time.
The detection methods based on CRISPR/Cas are widely used for detecting COVID-19, using crRNA combined with Cas12 or Cas13 to specifically target DNA or RNA to stimulate the trans-cleavage function of the Cas protein and so amplify the fluorescent signal of the reporter group to detect SARS-CoV-2 nucleic acid amplification products [14,20,21]. The amplification of viral nucleic acid is the key to whether it can be specifically detected. Comparative analysis of several isothermal amplification methods indicates that the primer design of the LAMP method is relatively complicated and that Standard RNA mix is serially diluted, and there were 7 replicates per dilution. e Visual OR-DETECTR assay by test card with lateral flow strip. Standard RNA mix is serially diluted, and there were 2 replicates per dilution it is hard to perform multiple LAMP amplifications [22]. Abbott ID Now, which is based on NEAR (Nicking Enzyme Amplification Reaction), requires only 5 min for a positive result. However, issues with false negativity have been raised due to its relatively high LoD. RT-RPA amplification has high sensitivity, and its primers are easily designed. Therefore, in our OR-DETECTR assay, we used RT-RPA as a potential means of template amplification. In this way, multiple RT-RPA can be performed.
We have developed two detection platforms, OR-DETECTR and OR-SHERLOCK. Through comparison, we found that: (1) the turnaround time, sensitivity, and specificity of OR-DETECTR based on Cas12a and OR-SHERLOCK based on Cas13a are approximately the same; (2) The design of crRNA is simple in the OR-SHERLOCK platform as the PAM region (Protospacer Adjacent Motif ) is not needed, but since Cas13a targets RNA, the detection mixture must be added to the components necessary for in vitro transcription, increasing the cost; (3) The crRNA used by OR-DETECTR requires a specific PAM region. As the genome of the SARS-CoV-2 virus is 29,903 nucleotides, it is easy to find suitable PAM regions. Additionally, OR-DETECTR targets DNA, meaning no further components are required.
According to current reports, most COVID-19 detection methods based on isothermal amplification and CRISPR/Cas require a two-step reaction. First, the reverse transcription and isothermal amplification are performed, after which a small amount of amplification product is taken for CRISPR/Cas detection. This is because isothermal amplification and the CRISPR detection interfere with each other. The transfer process requires opening the amplification reaction tube, which will cause aerosol pollution [13,14]. In clinical laboratories, if aerosol contamination occurs, false-positive results will be produced. Our OR-DETECTR test uses a one-tube detection method. We put the CRISPR mixture in the lid of the tube, and RT-RPA mixture in the bottom of the tube. After the RT-RPA reaction is completed, the amplification product is mixed with the CRISPR mixture by transient centrifugation, after which the fluorescence signal is detected. This reduces the operation of pipetting, shortens the time, and reduces the waste of consumables. Importantly, this can effectively avoid the unavailability of clinical laboratories caused by aerosol contamination.
Lateral flow paper has been applied to CRISPR-based detection methods for signal output. For single-sample detection, this is a good solution as no equipment is required to read the results, and it is promising for COVID-19 detection at home. According to current reports, its LoD is much lower than fluorescence-based detection methods [13]. Our OR-DETECTR assay can be easily read by a fluorescence plate reader to achieve sensitive and high-throughput SARS-CoV-2 detection. Additionally, the sensitivity of our lateral flow paper detection is the same as the fluorescence signal; that is, 2.5 copies per μl input. Finally, our OR-DETECTR test results are consistent with the rRT-PCR test results in the clinic, even for samples with very low viral loads. OR-DETECTR is suitable for clinical laboratories, pointof-care settings with the appropriate equipment, and perhaps for household purposes. Our OR-DETECTR is designed to provide a clear positive or negative result, rather than a quantitative viral load. Considering that the diagnosis of COVID-19 requires a fast, specific, and sensitive method, we do not think this is a disadvantage.