A Scalable, Easy-to-Deploy Protocol for Cas13-Based Detection of COVID19

There have been few end-to-end solutions to SARS-CoV-2 testing that fulfill the requirements of being immediately scalable and low in cost without sacrificing sensitivity. A scientist team, led by Dr. Maxwell Z. Wilson, University of California, is developing a CRISPR Cas-based SARS-CoV-2 RNA nucleic acid detection method that is low in cost, highly sensitive, and easy to deploy at sites with minimal infrastructure.

CRISPR Cas12 protein and CRISPR Cas13 protein based methods have been transformative with regard to pathogen detection. They are sensitive and, when coupled to isothermal amplification methods and lateral flow immunochromatography detection, have been made field deployable. These methods are promising tools for the detection of SARS-CoV-2. However, because of the global demand for testing, key reagents in these protocols are difficult to obtain. To lower the barrier to COVID-19 diagnostics, we devised a method we call CREST(Cas13-based, rugged, equitable, scalable testing). CREST addresses three ofthe main hurdles—reagent accessibility, equipment availability, and cost—that limit the scalability of Cas13-based testing, by taking advantage of widely available enzymes, low-cost thermocyclers, and easy-to-use fluorescent visualizers. Moreover, CREST is equivalent in sensitivity to the gold standard reverse transcription quantitative PCR (RT-qPCR) method most often deployed for COVID-19 testing. With these advantages, CREST has the potential to facilitate early detection of positive cases, regular monitoring of individuals at high risk, and implementation of informed containment measures for infected individuals.

METHODS AND RESULTS

To design a sensitive, low-cost, and easy-to-use SARS-CoV-2 nucleic acid detection method, we first identified critical steps that require limiting reagents, specific equipment, and highly trained individuals to perform them and thus present a barrier to testing at sites with limited infrastructure and resources. For CRISPR-Cas13-based methods, these steps include

(i) the amplification of the target material prior to detection and

(ii) the visualization of Cas13 activity by either colorimetric (immunochromatography) or fluorescent methods.

Fig.1. Overview of Cas13-based detection methods and CREST modifications. (i to iii) Standard sample collection, RNA extraction, and reverse transcription.

(iv) Amplification using cost-effective Taq polymerase and portable thermocyclers instead of isothermal reactions.

(v) Transcription and Cas13a activation are followed by fluorescence detection of dequenched poly(U) cleavage reporter visualized with a blue LED (∼495 nm) and orange filter or other fluorescence detection system.

To lower the first barrier, we analyzed options for detection of specific SARS-CoV-2 genomic sequences upon enzymatic amplification (Fig. 1). The gold standard method relies on RT-qPCR. Quantitative detection is accomplished using specialized instruments that detect fluorescent probes which report the extent of amplification of the target sequence in real time. While sensitivity is high (on the order of tens of target molecules per microliter), the main limitation is the requirement of real-time thermocyclers, analysis software, and trained personnel for data interpretation. Limitations aside, the core of the technology—amplification of a target nucleic acid sequence by PCR—is robust and sensitive and makes use of a widely available enzyme, Taq polymerase. These advantages motivated us to pair PCR with CRISPR-based detection of viral sequences, an approach that has been successful for the nucleic acid detection of DNA sequences using Cas9. The thermocyclers required for PCR are expensive, specialized instruments that are generally limited to professional laboratories. However, the recent “do-it-yourself biology”(DIYbio) movement has made PCR accessible through the creation of affordable, Bluetooth-enabled, field-ready thermocyclers, which can even be battery operated. These versatile thermocyclers can be used in unconventional environments and perform as well as traditional thermocyclers in moderate temperatures. We reasoned that these devices, such as the mini-PCR mini16, offer a low-cost solution for the amplification of the viral target material and can make COVID-19 testing widely available (Fig. 2).

Fig.2. Detection of SARS-CoV-2 RNA using CREST.

(A) The mini-PCR mini16 thermocycler and P51 molecular fluorescence visualizer used in this study. Both are portable, can be operated with batteries, and have minimal footprint.

(B) Fluorescence visualization of N1, N2, and N3 synthetic targets using the P51 visualizer.

To reduce the second hurdle, the visualization of test outcomes, we explored options available for the detection of CRISPR Cas13 activity. When bound to its target, CRISPR Cas13 catalyzes the nonspecific cleavage of RNAs. This target-specific recognition can be detected in many ways, for example, either by lateral flow immunochromatography or by fluorescence visualization through the use of a fluorescein- and quencher-conjugated poly(U) RNA cleavage reporter. Lateral-flow test strips area promising detection method. These strips utilize capillary action to move analytes through a solid support material striped with antibodies that can detect them, resulting in binary readouts. While they are simple to use and read—they may eventually enable in-home testing—their current availability islimited, and they are expensive relative to the cost per test (Fig. 3B; seealso Data Set S1 in the supplemental material). For this reason, we sought an affordable, scalable, easy-to-interpret, solution for visualization of positiveresults. In the CREST protocol, we use a P51 cardboard fluorescence visualizer, powered by a 9-V battery, for the detection of Cas13 activity instead of immunochromatography (Fig. 2).

Fig.3. Reagent cost/Test.

Comparative analysis of method sensitivity and reagent cost per test.

(A) (Left) Comparison of method sensitivity using aquantitative fluorescence detection instrument. (Right) RT-PCR + CRISPR Cas13a detection visualized with lateral flow strips.

(B) Associated costs of reagentsper test of each testing method (excluding up-front instrumentation costs). A test is defined as a single sample run in triplicate.

To validate a streamlined workflow with these devices, we measured the presence of the three viral sequences that correspond to the CDC RT-qPCR test (19). Briefly, we PAGE purified annealed synthetic DNA oligonucleotides flanked by an upstream T7 RNA polymerase promoter that encode sequences corresponding to the N1, N2, and N3 sites in the SARS-CoV-2 nucleocapsid gene. Next, we transcribed the DNA in vitro to obtain target RNAs. After CRISPR Cas13 purification and extensive optimization of the reaction conditions, we used these targets to determine the detection limit of the CREST protocol and found that we could detect as few as 10 copies of a target RNA molecule per μl (Fig. 2B). This result shows that CREST has sensitivity comparable to that of the corresponding RT-qPCR in our hands, demonstrating the power of CREST for pairing a thermal cycling amplification step (PCR) with a linear amplification step (transcription), combined with enzymatic signal amplification through fluorescence detection. In addition, we calculated CREST’s limit of detection in negative human nasopharyngeal (NP) swabs with spiked heat-inactivated SARS-CoV-2 virus (ATCCVR-1986HK) to be 200 copies per μl .

Next, we quantitively compared CREST’ssensitivity and cost to those of established methods. First, we compared it toRT-qPCR (one-step TaqMan assay) and found that, while being similarly sensitive, CREST’s reagents cost less than RT-qPCR’s even at the low scale of our pilot experiment (Fig. 3). In addition, the up-front cost of CREST instrumentation is 30 to 50 times lower. Second, we compared the RT-PCR amplification step of CREST to Cas13-based protocols that utilize RT-recombinase polymerase amplification (RT-RPA). We found thermal cycling amplification (20 cycles) to be substantially more efficient with comparable amplification reaction times (Fig. 3A). Moreover, in stark contrast tothe proprietary, high-cost, relatively small-batch production rates of reagents required for RPA, Taq polymerase, which has been in high-volume production for decades and is a workhorse of modern molecular biology, is readily accessible and stable at room temperature and lowers costs significantly (Fig. 3B). Last,we compared lateral-flow test strip visualization to CREST, and while we foundthem as sensitive as fluorescent detection methods (Fig. 3A), their high cost and difficulty to obtain can limit their distribution and scaled use in this pandemic.

To test the efficacy of CREST on human samples, we obtained 64 de-identified nasopharyngeal (NP) swabs from individuals collected through the Santa Barbara County Department of PublicHealth. We purified RNA from these samples, which were stored in viral transport medium, using a commercially available RNA extraction kit (QIAamp MiniElute virus spin kit; Qiagen). We used this RNA as input for a parallel comparison between CREST and the CDC-recommended one-step TaqMan assay (ourend-to-end CREST protocol is described in the supplemental material).Considering that CREST was designed to provide a binary outcome, we fit the CREST-to-TaqMan comparison to a sigmoid. We then calculated a goodness-of-fit Rvalue between assays for detection of N1, N2, and RNase P. These analyses revealed high concordance between CREST and TaqMan assays (R2 > 0.9for SARS-CoV-2 genes and R2 > 0.79 for RNase P). Of note, CREST appears to be more sensitive than TaqMan for detection of N1, whereas the converse appearsto be true for N2. We conducted two additional comparisons, one on 95 asymptomaticindividuals at University of California, Santa Barbara (UCSB), where oropharyngeal (OP) self-sampling was carried out, and a second one on 30 positive and 30 negative NP/OP samples obtained from University of California,San Francisco (UCSF). Taken together, the results show that CREST had a sensitivity of 97% and a specificity of 98%.

Finally, while we designed CREST to be an accessible and scalable assay for detecting SARS-CoV-2, it still requires RNA extraction using commercial kits, which limits its wide spread adoption. In acompanion paper, we present a method called PEARL (precipitation-enhanced analyte retrieval) which uses common laboratory reagents to bypass this limitation. To lift the final obstacle to CREST’s accessibility, we coupled PEARL with CREST and found that commercial RNA extraction could beomitted.

Source file

A Scalable, Easy-to-Deploy Protocol for Cas13-Based Detection of SARS-CoV-2 Genetic Material

Jennifer N. Rauch, Eric Valois, Sabrina C. Solley, Friederike Braig, Ryan S. Lach, Morgane Audouard, Jose Carlos Ponce-Rojas, Michael S. Costello, Naomi J. Baxter, Kenneth S. Kosik, Carolina Arias, Diego Acosta-Alvear, Maxwell Z. Wilson