Isothermal Amplification and Ambient Visualization in a Single Tube for the Detection of SARS-CoV-2 Using Loop-Mediated Amplification and CRISPR Technology.

We have developed a single-tube assay for SARS-CoV-2 in patient samples. This assay combined advantages of reverse transcription (RT) loop-mediated isothermal amplification (LAMP) with clustered regularly interspaced short palindromic repeats (CRISPRs) and the CRISPR-associated (Cas) enzyme Cas12a. Our assay is able to detect SARS-CoV-2 in a single tube within 40 min, requiring only a single temperature control (62 °C). The RT-LAMP reagents were added to the sample vial, while CRISPR Cas12a reagents were deposited onto the lid of the vial. After a half-hour RT-LAMP amplification, the tube was inverted and flicked to mix the detection reagents with the amplicon. The sequence-specific recognition of the amplicon by the CRISPR guide RNA and Cas12a enzyme improved specificity. Visible green fluorescence generated by the CRISPR Cas12a system was recorded using a smartphone camera. Analysis of 100 human respiratory swab samples for the N and/or E gene of SARS-CoV-2 produced 100% clinical specificity and no false positive. Analysis of 50 samples that were detected positive using reverse transcription quantitative polymerase chain reaction (RT-qPCR) resulted in an overall clinical sensitivity of 94%. Importantly, this included 20 samples that required 30-39 threshold cycles of RT-qPCR to achieve a positive detection. Integration of the exponential amplification ability of RT-LAMP and the sequence-specific processing by the CRISPR-Cas system into a molecular assay resulted in improvements in both analytical sensitivity and specificity. The single-tube assay is beneficial for future point-of-care applications.

Introduction

Effective control, containment, and mitigation of the coronavirus disease 2019 (COVID-19) require accessible tools and assays for timely diagnosis of the disease. The widely used molecular diagnostic assay relies on reverse transcription (RT) polymerase chain reaction (PCR), which enables the detection of specific genes in SARS-CoV-2, the virus that causes COVID-19.[1−4] Typically, specific sequences of any of the four genes that encode the RNA-dependent RNA polymerase (RdRp),[5,6] envelope (E),[7,8] nucleocapsid (N),[9,10] and spike (S)[11] proteins of SARS-CoV-2 are reverse-transcribed and exponentially amplified. The exponential amplification during PCR enables the detection of a few copies (molecules) of the gene sequence and results in the ultrasensitive detection of SARS-CoV-2 in infected individuals.

Global demands and competition for COVID-19 test kits and reagents necessitate the development of alternative assays and diagnostic tools. In addition, the need for thermal cycling for PCR tests makes point-of-care testing and on-site analysis in remote communities challenging. Consequently, there has been much interest in the development of molecular assays using isothermal amplification of nucleic acids.[12−17] Two isothermal amplification techniques, loop-mediated isothermal amplification (LAMP)[18−21] and recombinase polymerase amplification (RPA),[22−24] have been demonstrated to achieve the detection of SARS-CoV-2 RNA. LAMP is particularly promising for point-of-care applications because it requires only a single enzyme for the exponential amplification.[25]

Recent advances in clustered regularly interspaced short palindromic repeats (CRISPRs) and CRISPR-associated (Cas) proteins have stimulated the development of CRISPR technology and isothermal amplification techniques for molecular diagnosis of COVID-19. Broughton et al.[26] combined RT-LAMP with CRISPR Cas12a and developed a CRISPR-based assay for the detection of SARS-CoV-2. Named SARS-CoV-2 DNA endonuclease-targeted CRISPR trans reporter (DETECTR), this assay reduced the analysis time to <40 min, eliminated thermal cycling, and produced a color readout visible on lateral flow strips. The assay is potentially useful for point-of-care applications, although it required three steps: RT-LAMP amplification, CRISPR-mediated cleavage of reporters, and lateral flow display of results. The assay is conducted in two separate tubes under three temperature conditions: 62, 37 °C, and room temperature. Opening tubes to transfer reaction reagents and products increases the risk of potential contamination.

With the ultimate goal of simplifying the assay for point-of-care testing, we addressed the issues impeding the combination of RT-LAMP and CRISPR reactions in a single tube. We examined the apparent incompatibility between RT-LAMP and CRISPR, namely, differences in reaction temperatures and changes in the pH and Mg2+ concentration in the process of RT-LAMP. We report here an improved assay that incorporates RT-LAMP and CRISPR technology and demonstrates its application for the detection of SARS-CoV-2 in a single tube. In principle, the target sequence of viral RNA is reverse-transcribed and exponentially amplified by RT-LAMP and the specific amplicons are recognized by the CRISPR guide RNA and Cas12a enzyme. The integration of RT-LAMP with CRISPR takes advantage of both target amplification and sequence-specific recognition, resulting in improvements in both analytical sensitivity and specificity. Operationally, all necessary reactions are performed in a single closed tube, and the assay only needs a single temperature control (62 °C). The improvement achieved by reducing the number of steps and simplifying the assay requirement and thus eliminating the risk of contamination during the assay represents an important step toward point-of-care applications.

During the writing of this article, Joung et al.[27] reported “detection of SARS-CoV-2 with SHERLOCK one-pot testing”, built on the success of SHERLOCK (specific high-sensitivity enzymatic reporter unlocking).[28,29] The assay combined RT-LAMP with Cas12b and detected the N gene as the target. Recognizing that “LAMP operates at 55–70 °C and requires a thermostable Cas enzyme”, the authors generated and used Cas12b enzyme from Alicyclobacillus acidiphilus (AapCas12b) and identified the best combination of LAMP primers and guide RNAs to target the N gene. They tested “202 SARS-CoV-2-positive and 200 SARS-CoV-2–negative nasopharyngeal swab samples obtained from patients” and conclude that the assay “is suited for use in low-complexity clinical laboratories”.

The assay we report here targets both the E gene and N gene, while Joung et al.[27] detected the N gene. Our assay uses commercially available Cas12a enzyme, and this enzyme system requires only a short (41 nt) guide RNA (gRNA), whereas Cas12b used by Joung et al.[27] is not commercially available and it requires a longer (111 nt) single guide RNA (sgRNA) composed of both crRNA and tracrRNA.[30,31] The longer sgRNA could have a risk of “partial overlap between the sgRNA and one of the primers for LAMP”, which could “contribute to sporadic Cas12b collateral activity” and “occasional false-positive results”.[27] The 41 nt gRNA we use for the Cas12a system does not have sequence overlap with any of the LAMP primers.

Experimental Section

Viral RNA of SARS-CoV-2

The original SARS-CoV-2 virus strain (SARS-CoV-2/CANADA/VIDO 01/2020) was obtained from the University of Saskatchewan, Canada (a kind gift from Dr. Darryl Falzarano). SARS-CoV-2 was produced from the infection of Vero-E6 cells, and the amount of viral RNA was quantified using RT-qPCR for detecting the N gene (with N2 primers designed by the U.S. CDC).

Clinical Specimens and Extraction of RNA

One hundred respiratory swab specimens were collected by the Alberta Precision Laboratories (Canada). RNA was extracted using one of three platforms: the easyMAG (BioMerieux, Quebec, Canada) or the KingFisher Flex (Thermo Fisher Scientific) automated extraction and purification system or the STARlet automated extractor (Hamilton, Reno, NV). The specimen input was 200 μL, and the purified nucleic acid was extracted into 100 μL. An aliquot of 1–10 μL of the RNA extract was used as the sample input for the detection of the N gene and E gene of SARS-CoV-2.

Reverse Transcription Loop-Mediated Isothermal Amplification (RT-LAMP)

Sequences of the amplified regions of the N gene and E gene of SARS-CoV-2, along with sequences of the primer sets for RT-LAMP, are summarized in Supporting Information Table S1. All primers were synthesized by Integrated DNA Technologies (IDT, San Diego, CA).

Briefly, 25 μL of RT-LAMP reaction solution contained 1.4 mM deoxynucleotide (dNTP, NEB), 1× isothermal amplification buffer (NEB), 0.2 μM each of the outer primers (F3 and B3), 1.6 μM each of the inner primers (FIP and BIP), 0.8 μM each of the loop primers (LF and LB), 4 units of RNase inhibitor (Invitrogen), 7.5 units of WarmStart RTx reverse transcriptase (NEB), 8 units of Bst 2.0 DNA polymerase (NEB), and 1–10 μL of sample or nuclease-free water (as a negative control, NC). RT-LAMP reactions were performed at 62 °C for 30 min.

CRISPR Cas12a-Mediated Fluorescence Detection

The sequences of the guide RNAs (gRNAs), recognizing the specific sequences of the RT-LAMP amplicons, are listed in Table S2. Cas12a-mediated transcleavage of the ssDNA reporter, for the purpose of fluorescence detection, was carried out at room temperature (approximately 23 °C) for 10 min. The EnGen Lba Cas12a enzyme (NEB) at 1 μM concentration was preincubated with 1.25 μM gRNA in 1× NEBuffer 2.1 (NEB) to form the ribonucleoprotein (RNP) complex. An aliquot of the RNP complex solution was placed inside the cap of a PCR tube. The optimum concentrations of the Cas12a reaction reagents were 400 nM RNP complex, 40 mM MgSO4, and 10 μM ssDNA reporter in 50 mM Tris–HCl buffer (pH = 7.9).

RT-LAMP and CRISPR Cas12a-Mediated Ambient Visualization in a Single Tube for the Detection of SARS-CoV-2

A 25 μL mixture of the RT-LAMP reaction reagents was added to the bottom of a 0.2 mL PCR tube, and 10 μL of the Cas12a reaction solution was added inside the cap of the tube. An aliquot (2–10 μL) of the RNA extract sample was added to the bottom of the tube, mixing with the RT-LAMP reagents. The tube was then gently capped, preventing the Cas12a reagent from falling into the tube. The tube was placed in a temperature controller (heating block), and the bottom of the tube was maintained at 62 °C for 30 min to perform RT-LAMP. While the bottom of the tube was at 62 °C, the temperature in the cap was measured to be 31 °C. After 30 min of RT-LAMP, the tube was removed from the temperature controller, and the subsequent Cas12a-mediated detection step was performed at room temperature. Inverting and wrist-flicking the tube mixed the Cas12a reagents with the RT-LAMP amplicons. The tube was left at room temperature for 10 min. Green fluorescence was visualized under the excitation of a handheld UV lamp, and a photo was taken using a personal smartphone. The total time for the analysis of an RNA extract was 40 min, including 30 min for RT-LAMP and 10 min for Cas12a-mediated detection.

Assay Kit for RT-LAMP and CRISPR Cas12a-Mediated Ambient Visualization in a Single Tube

An assay kit was prepared to consist of a 0.2 mL PCR tube and a vial of the rehydration buffer solution containing primers and the reporter. The PCR tube was prepared to contain dried (lyophilized) reagents for the assay. The mixture of the RT-LAMP reaction reagents, except the primers, was added to the bottom of the PCR tube. This mixture was composed of 35 nmol dNTP, 1× isothermal amplification buffer, 150 nmol MgSO4, 4 units of RNase inhibitor, 7.5 units of WarmStart RTx reverse transcriptase, 8 units of Bst 2.0 DNA polymerase, and approximately 5 μmol d-(+)-trehalose dihydrate (Sigma). The RNP complex (22 pmol) for the Cas12a-mediated reaction, 400 nmol MgSO4, and 500 nmol Tris–HCl were added inside the cap of the tube. The tube was placed in a vacuum desiccator for 2 h to dry the reagents. After the reagents dried, the tube was capped and stored at 4 °C until use. The vial of rehydration buffer contained the primers for RT-LAMP and 5 μM ssDNA reporter for the Cas12a-mediated detection. The primers in the buffer solution can be customized according to the target of analysis.

Analysis of Clinical Specimens

Clinical samples were analyzed randomly in batches using the method of RT-LAMP amplification and CRISPR Cas12a-mediated detection, and the images resulting from the analysis were summarized in the order of their analysis (Table S3). Each batch of analyses included a positive control (PC) and a negative control. The positive control contained 2 μL or 10 μL of 750 copies/μL of viral RNA extracted from supernatants of Vero-E6 cell cultures. The negative control contained all of the reagents but no target, with the input sample being 2 or 10 μL of nuclease-free water. Samples were also analyzed using RT-PCR (Tables S4 and S5).

Results and Discussion

Principle and Operation of RT-LAMP Amplification and CRISPR-Mediated Visualization in a Single Tube for the Detection of SARS-CoV-2

The principle and operation of the assay that incorporates RT-LAMP amplification with CRISPR Cas12a-mediated detection are shown in Figure . Briefly, specific sequences of the E gene and N gene from SARS-CoV-2 (Table S1) are reverse-transcribed and amplified using RT-LAMP. The target sequence of the resulting amplicon is recognized by a predesigned guide RNA (gRNA) and interacts with the Cas12a–gRNA ribonucleoprotein (RNP) complex. Binding of the RNP to the specific amplicon activates the transcleavage activity of the Cas12a enzyme, resulting in cleavage of a short (8 nt) ssDNA reporter that is labeled at either end with a fluorophore/quencher pair. Cleavage of the reporter releases the quencher, allowing the fluorophore to fluoresce. The combination of RT-LAMP with CRISPR Cas12a takes advantage of the isothermal exponential amplification ability of RT-LAMP and the sequence recognition property and multiple turnover enzyme activity of the CRISPR Cas12a system, resulting in improvements in both sensitivity and specificity of the assay.

Figure 1

Schematics showing (A) general principle and (B) overall operation of the assay for the detection of SARS-CoV-2. (A) Specific gene sequence of SARS-CoV-2 RNA is amplified using RT-LAMP. The RT-LAMP products are scanned by the Cas12a–gRNA ribonucleoprotein (RNP) complex. The sequence in viral RNA marked in blue is the protospacer adjacent motif (PAM), which is essential for Cas12a recognition. The RNP binds to the specific sequence (in orange) complementary to gRNA, activating the transcleavage activity of Cas12a. The active Cas12a system cleaves a short ssDNA reporter (8 nt) that is labeled with a fluorophore and a quencher on either end. The cleavage of the reporter separates the quencher from the fluorophore, resulting in the generation of fluorescence. (B) PCR tube (0.2 mL) contains the RT-LAMP reagent mixture (25 μL or lyophilized) at the bottom and the Cas12a-mediated detection reagent mixture (10 μL of liquid or lyophilized) inside the cap. An aliquot (2–10 μL) of the RNA extract, and 25 μL of buffer if operated with lyophilized reagents, is added to the bottom of the tube, mixing with the RT-LAMP reagents. The tube is placed in a temperature controller, and the bottom of the tube is maintained at 62 °C for 30 min to allow for RT-LAMP reactions. The tube is then removed from the temperature controller, and the subsequent procedures are performed at room temperature. Inverting and wrist-flicking of the tube make the Cas12a reagents mix with the RT-LAMP amplicons in the bottom. Green fluorescence is generated at room temperature and is visualized under the excitation of a handheld UV lamp.

In practice, performing both RT-LAMP and CRISPR Cas12a reactions in a single tube is technically challenging because CRISPR Cas12a has been reported to perform at around 37 °C,[32,33] whereas the optimum reaction temperature for RT-LAMP is 60–65 °C.[34,35] To overcome the temperature incompatibility between the Cas12a-mediated reaction and RT-LAMP, we studied the effects of temperature and other conditions on RT-LAMP and CRISPR Cas12a-mediated reactions and developed a new approach of integrating RT-LAMP with CRISPR Cas12a in a single tube (Figure B). Briefly, 25 μL of mixture of reagents for RT-LAMP is placed at the bottom of a 0.2 mL PCR tube, and 10 μL of mixture of reagents for the CRISPR Cas12a-mediated detection is placed inside the cap of the PCR tube. The sample is added to the bottom of the tube, and the cap is carefully closed. The tube is placed in a heater where the bottom of the tube is maintained at 62 °C, the optimum temperature for the RT-LAMP reaction, whereas the temperature on the cap only reaches to about 31 °C (Figure S1). After 30 min of RT-LAMP, simple inverting and wrist-flicking the tube mix the Cas12a reagents with the amplicon generated by RT-LAMP. The CRISPR Cas12a-mediated recognition of the target and the cleavage of the reporter result in the generation of bright green fluorescence, visible with excitation by a handheld ultraviolet (UV) lamp (Figure S2). The same principle and operation apply when the assay is performed using the dehydrated reagent kit. The only difference is that the RT-LAMP primers and fluorescent reporter are in the buffer solution, and the remaining RT-LAMP reagents are dried at the bottom of the tube and the Cas12a–gRNA RNP is dried inside the cap.

This assay has three main advantages over other isothermal amplification assays. First, the CRISPR Cas12a-mediated detection improves the detection specificity. Second, the assay requires only a single controlled temperature (62 °C). Because isothermal amplification has a minimal requirement for temperature control, the assay is more amenable for future point-of-care applications. Third, the entire assay is performed in a single tube. After the addition of the sample, there is no need to open the tube, avoiding any cross-contamination of other samples by the amplicon.

Primers for RT-LAMP and gRNAs for Cas12a

We chose to detect both the E and N genes of SARS-CoV-2 because the E gene is highly conserved among all β coronaviruses, and the N gene can be used to differentiate SARS-CoV-2 from other coronaviruses. Our primer sets for RT-LAMP targeted the same gene sequences as those of the previously established RT-PCR assays for the E region (Charité Virology, Germany)[36] and for the N2 region (US CDC).[37] We used the gRNA for the E gene assay to possess a broad specificity for SARS-like coronaviruses, such as SARS-CoV-2, SARS-CoV, and bat SARS-like coronavirus (bat-SL-CoVZC45).[20] The gRNA for the N gene assay was designed to specifically recognize the N gene of SARS-CoV-2. Sequences of the primers and gRNAs are summarized in Tables S1 and S2.

Using RT-LAMP, we tested a series of solutions containing a range of concentrations of viral RNA, from 8 to 750 000 copies per μL. When amplification of RT-LAMP for the N and E genes was monitored in real time using SYBR Green detection, a plateau was reached before 30 min (Figure S3). To ensure sufficient amplification of RNA, we allowed 30 min for RT-LAMP.

Improving Specificity Using CRISPR Cas12a-Mediated Detection

To examine whether the CRISPR-mediated detection improved the specificity, we compared the real-time fluorescence detection using the SYBR Green dye with the CRISPR Cas12a-mediated detection of the RT-LAMP products (Figure ). When SYBR Green was used for detection, two of the triplicate negative controls also produced fluorescence within 30 min. These false-positive results from the negative controls are due to nonspecific amplification. However, when the CRISPR Cas12a system was used for the detection of the same RT-LAMP reaction products, there was no fluorescence from any of the negative controls (Figure ). These results show that the specificity is improved by incorporating the CRISPR-mediated detection with the RT-LAMP assay. Although the conventional SYBR Green detection can provide real-time monitoring of nucleic acids produced by RT-LAMP, nonspecific amplification products can potentially lead to false-positive results, especially in a resource limited setting. The incorporation of the CRISPR-mediated detection overcomes this problem by sequence-specific recognition of the amplicon. Binding of the predesigned gRNA to the complementary sequence of the target amplicon activates the Cas12a system to generate fluorescence signals. Nonspecific amplification products are not recognized by gRNA, and thus, the reporter is not cleaved to fluoresce.

Figure 2

Incorporating CRISPR Cas12a-mediated detection with RT-LAMP amplification improves the specificity. (A) RT-LAMP amplification curves from triplicate analyses of the target N gene and the negative controls, with the real-time fluorescence detection of intercalating SYBR Green. (B) CRISPR Cas12a-mediated detection of the RT-LAMP products. In this set of experiments, 25 μL of RT-LAMP reaction solution contained 1× NEBuffer 2.1 buffer, 1.4 mM deoxynucleotide (dNTP), 0.2 μM each of the outer primers (F3 and B3), 1.6 μM each of the inner primers (FIP and BIP), 0.8 μM each of the loop primers (LF and LB), 4 units of RNase inhibitor, 7.5 units of WarmStart RTx reverse transcriptase, 8 units of Bst 2.0 DNA polymerase, and 5 μL of 750 copies/μL viral RNA (as target) or nuclease-free water (as negative control). RT-LAMP reactions were performed at 62 °C, and the reaction products were monitored using either SYBR Green (A) or the CRISPR Cas12a system (B).

Issues of Different Temperatures Required for the RT-LAMP and Cas12a Reactions

Integrating both RT-LAMP and Cas12a reactions in a single tube would take advantage of both isothermal amplification by RT-LAMP and improved specificity of the Cas12a-mediated detection. We initially mixed all of the reagents, including primers, dNTP, reverse transcriptase, and polymerase for RT-LAMP with Cas12a, gRNA, and ssDNA reporter and performed reactions at different temperatures. We maintained the tubes at 62, 57, 52, 47, 42, 37, and 23 °C and monitored fluorescence. However, none of the experiments were successful. We reasoned that the main problem was the incompatible temperatures required for RT-LAMP and Cas12a reactions.

We tested the amplification of the N gene using RT-LAMP at different temperatures (62, 57, 52, 47, 42, and 37 °C) and monitored the generation of amplification products in real-time using SYBR Green (Figure S4). As expected, RT-LAMP carried out at 62 °C resulted in the most efficient amplification, needing less than 15 min to reach the exponential amplification phase. Although reactions at lower temperatures (57 and 52 °C) also generated products, the time needed to reach exponential amplification was prolonged: 20 min at 57 °C and more than 40 min at 52 °C. When the RT-LAMP was conducted at 47, 42, or 37 °C, no amplification was observed even after 60 min.

We also examined the ability of Cas12a to transcleave the ssDNA report at various temperatures: 62, 57, 52, 47, 42, 37, 31, and 23 °C (Figure ). These results indicate variable levels of the enzyme activity at 23–47 °C, peaking at 37 °C, and negligible activity above 52 °C. Therefore, the barrier to integrating RT-LAMP amplification with Cas12a-mediated detection was that the reaction temperature for RT-LAMP amplification (52–62 °C, Figure S4) was incompatible with Cas12a-mediated detection (23–47 °C, Figure ).

Figure 3

Fluorescence generated from the enzymatic cleavage of ssDNA reporters by Cas12a–gRNA maintained at different temperatures. Ten microliters of 50 nM of Cas12a–gRNA (recognizing the N gene) was maintained at the specified temperature for 30 min. Fifteen microliters of the mixture of DNA activator (for the N gene) and ssDNA reporter was then added to the tube. The final concentrations of Cas12a–gRNA, DNA activator, and ssDNA reporter were 20 nM, 1 nM, and 8 μM, respectively. Fluorescence was monitored every 1 min for 30 min, while the tube remained at the specified temperature. The inset shows net fluorescence (the difference between the fluorescence intensities at 30 min and at time 0).

Integrating RT-LAMP and Cas12a-Mediated Detection in a Single Tube

To overcome the problem of temperature incompatibility, we designed a two-step operation to combine the entire assay in a single tube (Figure B). We placed a mixture of the Cas12a enzyme, gRNA, and ssDNA reporter inside the cap of a common PCR tube. The temperature of the cap was approximately 31 °C (Figure S1) during the RT-LAMP reaction that took place at 62 °C at the bottom of the tube. Thus, the stability of the Cas12a reagents inside the cap was maintained.

We compared the Cas12a-mediated fluorescence detection at room temperature (approximate 23 °C) and 37 °C (Figure A). After 10 min, similar fluorescence signals were visible at both temperatures. Therefore, there is no need for a controlled temperature of 37 °C. Detection at room temperature simplifies the assay for potential on-site applications.

Figure 4

(A) Comparison of Cas12a-mediated fluorescence detection (10 min) at room temperature (approximate 23 °C) and 37 °C. (B) Cas12a-mediated fluorescence detection after different reaction times of the Cas12a-mediated cleavage of the reporter at room temperature. The RT-LAMP reaction was designed for targeting the N gene. The positive samples contained 3750 copies of SARS-CoV-2 RNA before the start of RT-LAMP. The negative controls contained all the reagents but no target; the input sample was nuclease-free water instead of the viral RNA.

We determined the reaction time needed for the Cas12a-mediated fluorescence detection at room temperature (Figure B). Fluorescence is clearly visible from the positive reactions as rapidly as 2 min after mixing the RT-LAMP product with the Cas12a reagents. There is no significant increase in fluorescence intensity beyond 10 min of Cas12a-mediated reactions. We subsequently chose 10 min of the Cas12a-mediated reaction time for the assay.

Optimization for RT-LAMP and Cas12a Reactions

Since the reagents are initially placed in two compartments of a single tube, an added benefit is the ability to fine-tune the RT-LAMP and Cas12a-mediated reaction conditions independently for compatibility and optimum performance. This optimization is particularly important because the available concentration of Mg2+ and the pH change as the RT-LAMP reaction progresses. We found that RT-LAMP decreased the concentration of Mg2+ to a level that was insufficient for the activity of Cas12a. Either the interaction of dNTP with Mg2+ decreased the concentration of Mg2+ available for the subsequent Cas12a-mediated reaction,[38] or pyrophosphate generated during RT-LAMP could form magnesium phosphate precipitates,[39] decreasing the concentration of Mg2+ available for the subsequent Cas12a-mediated reaction. The addition of Mg2+ to the Cas12a reaction mixture could regain the activity of Cas12a (Figure S5). We therefore added Mg2+ into the cap containing the Cas12a reagent mixture to compensate for the loss of Mg2+ during the RT-LAMP reaction. We found that 40 mM Mg2+ in the Cas12a reagent mixture was sufficient to compensate for the loss of Mg2+ and ensure the activity of Cas12a (Figure S6).

Protons liberated during RT-LAMP reactions decreased the pH of the reaction mixture from 8.8 at the beginning of the reaction to 6.0 at the end of RT-LAMP.[40] To ensure the activity of Cas12a in the subsequent reactions, we compensated for changes in pH. Placing the Cas12a-mediated reaction reagents separate from the RT-LAMP reagents facilitated pre-adjustment of the pH in the Cas12a reagent mixture to compensate for changes that occurred during the RT-LAMP process. We determined that 50 mM Tris–HCl buffer (pH 7.9) in the Cas12a reagent mixture was sufficient to compensate for the pH changes during RT-LAMP and to achieve the optimum pH condition for the transcleavage activity of the Cas12a system.

Any unused primers after RT-LAMP can compete with the ssDNA reporter and serve as substrates for the transcleavage activity of Cas12a. Therefore, we optimized the concentration of ssDNA reporter in the Cas12a reagent to reduce the effect of unused primers and to ensure the generation of sufficient fluorescence for visualization. We found that 10 μM reporter was optimum for achieving the brightest fluorescence with low background (Figure S7).

Limit of Detection and Reproducibility

We detected the N gene and E gene of SARS-CoV-2 at a wide range of RNA concentrations, from 0 to 750 000 copies/μL. SARS-CoV-2 RNA was extracted from supernatants of Vero-E6 cell cultures, and the concentrations of SARS-CoV-2 RNA were measured using RT-qPCR targeting the N gene. Samples containing viral RNA concentrations 30 copies/μL or higher consistently gave positive results from the detection of the N gene (Figure ). Samples that contained 1 copy/μL (5 copies for 5 μL sample input) of the target gave occasional positive results (3 out of 12) (Figure S8); however, consistently accurate detection was achieved when the concentration was above 30 copies/μL (150 copies total). True positive rates from 12 replicate analyses were 100% (30 copies/μL), 83% (15 copies/μL), 67% (10 copies/μL), 25% (5 copies/μL), and 25% (1 copy/μL) (Figure S8 and Table S6). Random sampling errors from the analysis of small numbers of molecules probably contribute to the observed detection rates. Similar random sampling errors also occur in other assays, e.g., RT-qPCR (Table S7). We report a detection limit of 30 copies/μL (150 copies), for a consistent 100% true positive rate. Similar results were obtained from the detection of the E gene, where consistently positive results were obtained when the concentration was higher than 45 copies/μL (Figure S9).

Figure 5

Representative images obtained from the detection of the N gene at a range of concentrations (0, 8, 15, 30, 45, 60, 75, 750, 7500, 75 000, and 750 000 copies/μL). Five microliters of sample was used.

We determined the reproducibility by conducting 12 replicate analyses of the N gene in a sample that contained 750 copies/μL of SARS-CoV-2 RNA and a negative control (nuclease-free water) (Figure S10). Relative standard deviation (RSD) was 6–7% on the basis of fluorescence intensity measurements and 3–10% from measurements of digitized color intensity.

Analysis of Clinical Specimens

We carried out our assay on 100 clinical samples, consisting of 50 clinical negative and 50 SARS-CoV-2 positive samples. Results from the analysis of multiple batches of clinical samples on multiple days are summarized in Table S3. Overall, we accurately assessed 50 true negatives, 47 true positives, and 3 false negatives from a total of 100 clinical samples. The three false-negative samples contained very low concentrations of the target RNA, requiring 33–39 threshold cycles of RT-qPCR to achieve a positive detection. These results demonstrate an overall clinical specificity of 100% (detecting all 50 negatives) and an overall clinical sensitivity of 94% (detecting 47 positives of 50 total positive samples).

The 50 positive samples contained a wide range of viral RNA concentrations, according to the RT-qPCR analyses of the E gene (Ct ranging from 11.7 to 39.2). Of 30 samples that had Ct values below 30, all were positively detected using both the N and E genes, equivalent to a clinical sensitivity of 100% (detecting all 30 positives). Our assay was also successful in correctly detecting 17/20 positives that required more than 30 cycles to be positively detected by RT-qPCR.

Analyses of all 50 negative samples for both the N gene and E gene consistently yielded negative results. To ensure that our detection of the negative signal was not due to insufficient samples, we increased the sample amount from 2 to 5 μL and repeated the analysis. Replicate analyses of 6 representative samples still resulted in no signal from these true negative samples (Figure S11). There was no false positive from the analysis of any of the 100 samples, confirming the 100% clinical specificity of the assay.

Analysis Using a Single-Tube Assay Kit

To make the assay kit convenient for shipping and to demonstrate its potential for on-site applications, we have prepared PCR tubes containing dried reagent mixtures and conducted the assay after rehydrating the reagents. The main benefit is that the prepared tubes can be easily shipped to the field and used conveniently to conduct the assay. Our results from the analysis of controls (Figure A) and representative samples (Figure B) confirmed the successful detection of SARS-CoV-2 using the prepared tube and reagent kit.

Figure 6

Analysis of the controls (A) and human samples (B) using the prepared assay kit containing dehydrated reagents. The Cas12a reagent mixture containing the RNP complex, MgSO4, and Tris–HCl was dehydrated on the cap. The RT-LAMP reagent mixture (except primers) was dehydrated at the bottom of the tube. (A) Positive control contained 5 μL of 7500 copies/μL viral RNA as the input sample. The negative control contained 5 μL of water as the sample input. (B) Sample #29 was confirmed negative and sample #72 was confirmed positive using RT-PCR. The positive control (PC) contained 2 μL of 7500 copies/μL of viral RNA. The negative control (NC) contained all of the reagents but no target, with the input sample being 2 μL of nuclease-free water.

Conclusions

Our assay for the detection of SARS-CoV-2 has two appealing features: (1) conducting CRISPR-Cas detection at room temperature, thus only requiring a single controlled temperature for isothermal amplification, and (2) integration of isothermal amplification and subsequent CRISPR-Cas detection in a single tube, which simplifies the operation and eliminates the risk of contamination during the assay. These features are a significant advance in the application of isothermal amplification techniques to point-of-care applications and on-site analysis. The RT-LAMP-Cas12a assay is not limited to the detection of SARS-CoV-2. We envision the application of the assay for other infectious agents through simply altering primers and gRNA to target other nucleic acids.