Digital CRISPR/Cas‐Assisted Assay for Rapid and Sensitive Detection of SARS‐CoV‐2

Abstract The unprecedented demand for rapid diagnostics in response to the COVID‐19 pandemic has brought the spotlight onto clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR‐associated systems (Cas)‐assisted nucleic acid detection assays. Already benefitting from an elegant detection mechanism, fast assay time, and low reaction temperature, these assays can be further advanced via integration with powerful, digital‐based detection. Thus motivated, the first digital CRISPR/Cas‐assisted assay—coined digitization‐enhanced CRISPR/Cas‐assisted one‐pot virus detection (deCOViD)—is developed and applied toward SARS‐CoV‐2 detection. deCOViD is realized through tuning and discretizing a one‐step, fluorescence‐based, CRISPR/Cas12a‐assisted reverse transcription recombinase polymerase amplification assay into sub‐nanoliter reaction wells within commercially available microfluidic digital chips. The uniformly elevated digital concentrations enable deCOViD to achieve qualitative detection in <15 min and quantitative detection in 30 min with high signal‐to‐background ratio, broad dynamic range, and high sensitivity—down to 1 genome equivalent (GE) µL−1 of SARS‐CoV‐2 RNA and 20 GE µL−1 of heat‐inactivated SARS‐CoV‐2, which outstrips its benchtop‐based counterpart and represents one of the fastest and most sensitive CRISPR/Cas‐assisted SARS‐CoV‐2 detection to date. Moreover, deCOViD can detect RNA extracts from clinical samples. Taken together, deCOViD opens a new avenue for advancing CRISPR/Cas‐assisted assays and combating the COVID‐19 pandemic and beyond.


Bulk
Cas12a-assisted RT-RPA Bulk Cas12a-assisted RT-RPA was performed in a Bio-Rad CFX96 Touch Real-Time PCR Detection System (Bio-Rad, Hercules, CA). Assembled Cas12a-assisted RT-RPA mix was pipetted into PCR strips (Bio-Rad, Hercules, CA) at a final volume of 10 L for performing bulk reactions. All bulk reactions were incubated at 42 °C (or the temperatures indicated in Fig. S1) for 60 min, and the fluorescence signals were measured every 1 min. The fluorescence signals measured by the Bio-Rad CFX96 system were displayed without baseline subtraction (i.e., under "No Baseline Subtraction" mode in the CFX Manager Software). A saturated fluorescence intensity was the maximum intensity which the Real-Time PCR Detection System could determine.
2.3. Loading Cas12a-assisted RT-RPA mix into QuantStudio digital chip To perform digital Cas12a-assisted RT-RPA (deCOViD), assembled Cas12a-assisted RT-RPA mix was first loaded into a commercial QuantStudio 3D Digital PCR 20K Chip v2 (ThermoFisher Scientific, Waltham, MA) by using a QuantStudio 3D Digital PCR Chip Loader and closely following the user guide. Each QuantStudio 3D Digital PCR 20K Chip v2 has a thermally conductive case that securely houses a 10 mm × 10 mm high-density reaction plate containing 20000 reaction wells on its surface; each reaction well is ~700 picoliter in volume for discretizing Cas12a-assisted RT-RPA into independent digital reactions. The loading process began by installing a QuantStudio 3D Digital PCR 20K Chip v2, a QuantStudio 3D Digital PCR Chip Lid v2 (the corresponding adhesive lid with optical window), and a QuantStudio 3D Digital PCR Sample Loading Blade in their respective positions of the loader. Next, 15 µL Cas12aassisted RT-RPA mix was pipetted in the sample loading port of the loading blade. The loader then automatically moved the loading blade across the chip, simultaneously dispensing the reaction mix through the loading blade into the reaction wells. Subsequently, several drops of Immersion Fluid were gently (without touching the chip surface) to cover the entire chip. Then, by rotating the loader arm, the chip lid was brought into contact with the chip case and firmly pressed into a tightly sealed assembly. Additional Immersion Fluid was then gently dispensed via a syringe through the fill port in the chip lid while holding the chip and lid assembly at a ~45° angle and allowing air to escape from the fill port, until the chip case contained only an air bubble < 2 -3 mm in diameter. Finally, the label on the chip lid was firmly pressed over the fill port to establish a fully sealed chip with digitized reaction mix ready for performing digital Cas12a-assisted RT-RPA (deCOViD)either in end-point format or real-time format.
2.4. End-point digital Cas12a-assisted RT-RPA (deCOViD) End-point digital Cas12a-assisted RT-RPA (deCOViD) was performed by first heating chips at 42 °C on a ProFlex 2× flat PCR System (Thermo Fisher Scientific, Waltham, MA) and, after heating, measuring the fluorescence signals from the chips via fluorescence microscopy. For heating, each chip was placed within a grid of QuantStudio 3D Digital PCR Chip Adapters that were placed on the sample block of the ProFlex system, which was set to 42 °C. QuantStudio 3D Digital PCR Thermal Pads were laid over the chip adapters before closing the lid of the Proflex system and commencing the reaction. All reactions were performed for either 60 min (e.g., Figure 2) or 30 min (e.g., Figure 3). Of note, we followed the user guide and elevated the chips in the Proflex system during digital Cas12a-assisted RT-RPA (deCOViD), though such an elevation was likely ineffectual to digital Cas12a-assisted RT-RPA (deCOViD). Specifically, the chips were positioned such that their fill ports orient toward the front of the ProFlex system, which was raised to an 11° incline by the underneath QuantStudio 3D Tilt Base. The elevated chips were meant to allow air bubbles formed during thermocycling to float to the elevated fill ports and away from the reaction wells, thus ensuring successful digital PCR. On the other hand, the low, 42 °C reaction temperature of digital Cas12a-assisted RT-RPA (deCOViD) likely minimized air bubbles and thus rendered chip elevation ineffectual.
After heating, each chip was taken from the Proflex system to a fluorescence microscope to measure fluorescence signals from digital Cas12a-assisted RT-RPA (deCOViD). The fluorescence microscope (BX51, Olympus, Japan) is equipped with a collimated LED light source (M625L4-C1; Thorlabs, Inc., Newton, NJ), an Alexa647-compatible filter cube (Semrock Cy5-4040C-OMF; IDEX Health & Science, LLC, Rochester, NY), a 4× magnification objective lens (Olympus UPlanFL N 4×/0.13 NA), and a digital CCD camera (Retiga EXi Fast 1394, QImaging, Canada). The Alexa647-compatible filter cube has a 628 ± 20 nm bandpass excitation filter, a 660 nm dichroic beamsplitter, and a 692 ± 20 nm bandpass emission filter. The CCD camera is connected to a PC and interfaced with QCapture software (QImaging, Canada). For each chip, fluorescence micrographs of 15 distinct regions were taken to cover ~12000 reaction wells. At each region, the shutter was manually opened, a 12-bit fluorescence micrograph was immediately taken under 50 ms exposure, and the shutter was quickly manually closed.

Real-time digital Cas12a-assisted RT-RPA (deCOViD)
Real-time digital Cas12a-assisted RT-RPA (deCOViD) was performed by heating each chip at 42 °C on a custom heater and concurrently measuring the fluorescence signals from the chips via fluorescence microscopy. The custom heater was made to fit between the fluorescence microscope stage and the 4× magnification objective lens. This was accomplished by assembling the heater with a 50 mm (length) × 50 mm (width) × 3.8 mm (height) standard Peltier module (Custom Thermoelectric, LLC, Bishopville, MD) and a 50 mm (length) × 50 mm (width) × 6 mm (height) heatsink (ATS-CPX050050006-199-C2-R0, Digi-Key, Thief River Falls, MN). The temperature of the heater was controlled by a FTC100D Controller (Accuthermo Technology Corp., Fremont, CA). After the chip was placed on the custom heater and the region to be detected in the chip was selected, the heater was raised to 42 °C and the same region was fluorescently imaged every 1 min for 60 min. At every minute, the shutter was manually opened, a 12-bit fluorescence micrograph was immediately taken under 50 ms exposure, and the shutter was quickly manually closed.
2.6. End-point digital RT-PCR End-point digital RT-PCR was performed by first assembling an in-house RT-PCR assay using the US CDC-approved SARS-CoV-2 N1 and N2 primers and probes [2] . The 15 L assay mix contained 1× qScript XLT 1-Step RT-qPCR ToughMix, 500 nM each of CDC N1 primer, 250 nM Cyanine 5 (Cy5)-labeled CDC N1 DNA probe, 750 nM each of CDC N2 primer, 250 nM Fluorescein amidites (FAM)-labeled CDC N2 DNA probe, 1 mg/mL BSA, 0.1% Tween-20, and synthetic SARS-CoV-2 RNA (at varying concentrations). Assembled RT-PCR mix was then loaded into a commercial QuantStudio 3D Digital PCR 20K Chip v2 as described above. The reverse transcription reaction was initiated by chip incubation at 50 °C for 2 min (reverse transcription), followed by PCR amplification at 95 °C for 1 min (hot start), and 50 cycles of 95 °C for 5 s, and 50 °C for 20 s on a ProFlex 2× flat PCR System. After PCR amplification, digital chips were taken from the Proflex system to a fluorescence microscope to measure fluorescence signals from digital RT-PCR as described above. Of note, for consistency with deCOViD, only the fluorescence signals from Cy5-labeled CDC N1 DNA probe was detected by using the same excitation and emission filters as deCOViD.

Clinical sample testing
Four de-identified clinical nasopharyngeal swabs (including 2 COVID-19 positive samples) in universal transport medium were obtained from the Johns Hopkins Hospital Clinical Microbiology Laboratory in compliance with ethical regulations and the approval of Institutional Review Board (IRB00246027). Viral RNAs from these 4 samples were extracted via ChargeSwitch magnetic beads (Thermo Fisher Scientific, Waltham, MA) following the manufacturer's instructions. Ten μL of each sample was used as the input for extraction and eluted with 20 μL elution buffer. Prior to testing with the bulk CRISPR/Cas-assisted assay and deCOViD, eluted RNAs from these samples were screened for SARS-CoV-2 by the in-house RT-PCR assay (with qScript XLT 1-Step RT-qPCR ToughMix and US CDC-approved SARS-CoV-2 primers and probes) performed in a Bio-Rad CFX96 Touch Real-Time PCR Detection System under real-time detection mode. RT-qPCR is performed at 50 °C for 2 min (reverse transcription), followed by 95 °C for 1 min (hot start) and 50 cycles of 95 °C for 5 s and 50 °C for 20 s, with the fluorescence measured at 50 °C of each cycle.
2.8. Data processing and statistical analysis Data acquisition from fluorescence micrographs and subsequent data analysis were performed using ImageJ (1.52p), MATLAB (R2019b), Microsoft Excel 365, and Origin 2018. Region of interest (ROI) from each fluorescence micrograph was defined with ImageJ. Fluorescence intensities from the reaction wells in the QuantStudio 3D Digital PCR Chip were measured with ImageJ and MATLAB. Downstream data analyses including statistical analysis, plotting, and data fitting were performed with Excel and Origin.
For end-point digital analysis, each fluorescence micrograph was first converted to 8-bit grayscale in ImageJ. Using the publicly available "AdaptiveThreshold" plugin function in ImageJ, the "Mask" image was created. Subsequently, the ROIs from each "Mask" image were defined using the "Analyze particles" function in ImageJ and saved for downstream fluorescence intensity measurements. Each ROI file from ImageJ was then converted into a MATLAB structure using the publicly available MATLAB program, "ReadImageJROI". A custom-built MATLAB program was developed for fluorescence intensity measurements using fluorescence micrographs, the "Mask" images, and corresponding ROIs. The program measured fluorescence intensities of each reaction well from each of 15 micrographs and counted the number of reaction wells with the fluorescence intensity below and above threshold for quantifying negative and positive wells, respectively. The threshold value was empirically set at 15000 and sufficiently higher than non-specific signals from no-template controls, which could sometimes reach ~10000. The program then combined 15 measurements for calculating the percent positive (i.e., the number of positive reaction wells divided by the total number of reaction wells) of each digital reaction and extracting the end-point fluorescence intensities of positive wells.
For real-time digital analysis, a single ROI was defined from a micrograph at 1 min as described above. With a single ROI, fluorescence intensities of reaction wells in micrographs at each min were measured using "Multi Measure" function in ROI manager. The measurements were exported, and real-time curves from individual reaction wells were plotted via Origin. The timeto-positive values were then calculated via custom-built MATLAB program for each positive well, a time point where fluorescence intensity first exceeds the threshold.
All downstream data analyses were performed via Excel and Origin. The time-to-positive (i.e., Figure 2e) and the signal-to-background ratio (i.e., Figure 2f) from each RNA concentration were calculated from triplicated results of bulk experiment (i.e., n = 3) and collected positive signals (n > 3) from a single digital chip, and presented as mean ± SD. For the signal-tobackground ratio calculation, the end-point fluorescence intensity at 30 min (F) was divided by the initial fluorescence intensity at 1 min (F initial ) for both bulk and digital Cas12a-assisted RT-RPA reaction. The calculated F/F initial values from each synthetic SARS-CoV-2 RNA concentration were plotted as a scatter plot. For sensitivity comparison between bulk and digital Cas12a-assisted RT-RPA reaction with both synthetic SARS-CoV-2 RNA (i.e., Figure 3a) or heat-inactivated SARS-CoV-2 (i.e., Figure 3b), the end-point fluorescence intensities at 30 min in bulk reaction and the percent positives in digital reaction were plotted, and linearly fitted via Origin. All RNA concentrations were tested in triplicate (i.e., n = 3) with both bulk reaction and digital reaction. The data are presened as mean ± SD. Table S1. Sequences of RPA primers, Cas12a-guide RNAs, and ssDNA reporter.