One-Pot Visual Detection of African Swine Fever Virus Using CRISPR-Cas12a

African swine fever virus (ASFV) is a leading cause of worldwide agricultural loss. ASFV is a highly contagious and lethal disease for both domestic and wild pigs, which has brought enormous economic losses to a number of countries. Conventional methods, such as general polymerase chain reaction and isothermal amplification, are time-consuming, instrument-dependent, and unsatisfactorily accurate. Therefore, rapid, sensitive, and field-deployable detection of ASFV is important for disease surveillance and control. Herein, we created a one-pot visual detection system for ASFV with CRISPR/Cas12a technology combined with LAMP or RPA. A mineral oil sealing strategy was adopted to mitigate sample cross-contamination between parallel vials during high-throughput testing. Furthermore, the blue fluorescence signal produced by ssDNA reporter could be observed by the naked eye without any dedicated instrument. For CRISPR-RPA system, detection could be completed within 40 min with advantageous sensitivity. While CRISPR-LAMP system could complete it within 60 min with a high sensitivity of 5.8 × 102 copies/μl. Furthermore, we verified such detection platforms display no cross-reactivity with other porcine DNA or RNA viruses. Both CRISPR-RPA and CRISPR-LAMP systems permit highly rapid, sensitive, specific, and low-cost Cas12a-mediated visual diagnostic of ASFV for point-of-care testing (POCT) applications.


INTRODUCTION
African swine fever (ASF) is a highly lethal and contagious disease in domestic pigs caused by the African swine fever virus (ASFV) and is a notifiable disease by the World Organization for Animal Health (OIE) (1). ASFV is a large and complex double-stranded DNA arbovirus that is the only member of Asfivirus genus (2). Based on the highly conserved gene B646L encoding the viral protein p72, ASFV is currently classified into 24 genotypes. In 2018, ASF began to spread in China, where the virus circulating was identified as genotype II (1,3), an epidemic strain that was 100% consistent with that in Russia (4). Due to the absence of effective treatments or vaccines, ASF disease control mainly relies on culling pigs (2,5). With high infectivity and mortality, ASFV has seriously affected animal husbandry.
The focus of prevention and control of ASF is currently still in the early diagnosis and outbreak control stages. Therefore, accurate and efficient laboratory diagnosis is of vital importance.
Quantitative PCR (qPCR) and conventional PCR, which are also recommended by OIE (6)(7)(8), are sensitive methods for the detection of ASFV. However, the dependence on expensive thermocyclers and skilled operators limits the application of these methods for point-of-care (POC) detection (9,10). Isothermal amplification techniques, such as recombinase polymerase amplification (RPA) (11)(12)(13), loop-mediated isothermal amplification (LAMP) (14,15), polymerase crosslinking spiral reaction (PCLSR) (16), and cross-priming amplification (CPA) (17), have been suggested for the detection of ASFV based on the highly conserved region of essential ASFV genes, such as p72. However, the reaction temperature close to room temperature can easily generate false-positive test results (18,19). Therefore, it is necessary to develop a sensitive, specific, equipment-free, and visual method for the detection of ASFV.

Preparation of Genomic DNA Samples
The 1941bp fragment (Genomic Sequence: NC_001659.2) of the ASFV p72 gene (also known as B646L) was chemically synthesized and cloned into pUC57 plasmid (herein referred to as pUC57-p72 DNA) by Sangon Biotech (Shanghai). Co. Ltd. The pUC57-p72 DNA was used as the template for the optimization of the detection system, as well as for the determination of sensitivity, as standard ASFV plasmid was used in previous reports (16,17). The DNA or cDNA of the pseudorabies virus (PRV), porcine reproductive and respiratory syndrome virus (PRRSV), porcine epidemic diarrhea virus (PEDV), and porcine deltacoronavirus (PDCoV) were obtained from the Shanghai Veterinary Research Institute (Chinese Academy of Agricultural Sciences), for the use as samples for specificity determination.

Oligonucleotide Primers for Amplification and crRNA Preparation
The most conserved region of the gene was subjected to design isothermal amplification primers. The RPA primers were designed using online software (Primer-blast) according to the TwistAmp assay. These forward and reverse primers formed a number of primer pairs, and the RPA products should be 100-200 bp. The LAMP primers were designed using PrimerExplorer V5 (http://primerexplorer.jp/e/), comprising of two outer primers F3 and B3, two inner primers FIP and BIP, and two loop primers LF and LB.
Using CHOPCHOP (https://chopchop.cbu.uib.no/), the 23nt crRNA targets were designed targeting the p72 gene, which were also the targets of the RPA and LAMP. The primers and crRNAs sequences, listed in Table 1, were synthesized by Sangon Biotech.

Establishment and Optimization of the One-Pot Visual Detection System
One-pot detection combines isothermal pre-amplification and CRISPR/Cas12a-mediated cleavage detection in the same reaction tube. Briefly, the RPA or LAMP pre-amplification assays were added to an Eppendorf tube, and 35 µl of mineral oil was covered on the pre-amplification assay. After the RPA or LAMP reaction, 20 µl of CRISPR reaction buffer (pre-added inside the lid) was mixed with 25 µl of the amplification assay by hand shaking or spinning down in a minifuge for 5 s. The tube was put in PTC-200 thermocyclers at 37 • C for 20 min, and the fluorescence signal after the CRISPR/Cas12a-mediated cleavage reaction was visualized using a transilluminator under blue light (Figure 1). For the Cas12a-RPA one-pot detection system, a CRISPR reaction optimization experiment (incubated at 37 • C for 5, 10, 20, and 30 min) was conducted to find an appropriate reaction time to shorten the one-pot detection. In addition, with a transfer step in between, the RPA and CRISPR reactions can be run sequentially. To optimize the one-pot detection, a single combined mixture of RPA and CRISPR reactions was carried out at the same reaction temperature (37 • C), which was "one-step" in one-pot detection. After 35 µl of mineral oil was used to cover the surface of the combined mixture, the reaction tube was put in PTC-200 thermocyclers at 37 • C for 20, 30, 40, and 50 min to optimize the detection system.
The reaction time of the CRISPR-LAMP one-pot detection system was optimized separately for different temperatures (65 • C for LAMP and 37 • C for CRISPR reaction). Briefly, LAMP preamplification was incubated at 65 • C for 20, 30, 40, and 50 min, and CRISPR/Cas12a-mediated cleavage reaction was incubated at 37 • C for 5, 10, 15, 20, 25, and 30 min after an optimized LAMP reaction time in one tube.

Evaluation of the One-Pot Visual Detection System
The specificity of the one-pot visual detection systems was determined by the optimized procedure, the genomic DNA or cDNA of PRV, PRRSV, PEDV, and PDCoV. Further, pUC57-p72 DNA and negative controls (ddH 2 O) were also used, with the amount of genomic DNA or cDNA being 1 µl per reaction. In addition, all the genomic DNA or cDNA were used as templates for the PCR with their own specific primers (Supplementary Table 1) before evaluating the specificity of the one-pot visual detection system.
The analytic sensitivities of the newly developed one-pot visual detection systems were determined with the pUC57-p72 DNA ranging from 7 × 10 9 to 7 × 10 0 copies/µl using multiple dilution methods. The reaction mixtures were heated at 37 • C or 65 • C for an optimal amount of time. The ASFV B646L gene plasmid reference material [GBW(E) 091034], with high stability and uniformity, as well as an extended uncertainty of 0.9 × 10 3 copies/µl (K = 2), was used for further determination of the sensitivity of both the CRISPR-RPA and CRISPR-LAMP one-pot detection systems. The reference material was diluted, ranging from 5.8 × 104 to 5.8 × 10 0 copies/µl, and detected as previously described in the method section.
For further evaluation, diluted pUC57-p72 DNA was used as a template to compare the newly developed one-pot

Establishing the One-Pot Detection Assay
By adding isothermal pre-amplification and Cas12a-mediated cleavage reaction together in one tube, with mineral oil covering the surface of the LAMP/RPA amplification reagent, a onetube visual detection system was assembled. For optimizing the CRISPR-RPA one-pot detection, the result (Figure 2A) showed that the fluorescence signal increased rapidly with time until it reached the peak value while the negative control remained no fluorescence signal. Moreover, the fluorescence signal was clear enough for the naked eye to detect at 20 min. Therefore, a total of 40 min (RPA for 20 min and CRISPR reaction for 20 min) reaction time were established for the CRISPR-RPA onepot detection. Besides, after 40 min reaction time, the "onestep" in one-pot detection also showed an increased pattern FIGURE 1 | One-pot visual detection system. One-pot visual detection system showing the process of DNA pre-amplification, Cas12a/crRNA cleavage reaction, and fluorescence visualization. RPA or LAMP reagent, crRNA, Cas12a enzyme, and ssDNA-FQ reporter are all in one tube, with mineral oil sealing. Target binding of Cas12a will unleash its ability to digest ssDNA-FQ reporter. The fluorophore FAM (F) is quenched by a quencher (Q) if intact and emits fluorescence when cleaved, which is visual under the blue light.
in the same fluorescence signal until reaching the peak value (Supplementary Figure 1).
For the CRISPR-LAMP one-pot detection system, the optimization was separated into two parts, LAMP optimization and CRISPR reaction optimization. After incubating at 65 • C for different durations and a Cas12a-mediated reaction for 40 min, the result ( Figure 2B) showed that 40 min is an appropriate pre-amplification time for LAMP. After LAMP pre-amplifying for the optimized time (40 min), a CRISPR reaction was conducted to adjust the time. The result ( Figure 2C) showed that 20 min is an optimized time for CRISPR reaction. Eventually, a total of 60 min (40 min for LAMP and 20 min for CRISPR reaction) reaction time were established for CRISPR-LAMP one-pot detection.

Specificity of the One-Pot Visual Detection System
The specificity of the one-pot visual detection system was evaluated using pUC57-p72 DNA and other porcine viruses' genomic DNA or cDNA, including PRV, PRRSV, PEDV, and PDCoV. The specific of all the genomic DNA or cDNA were verified with PCR and the results of which are displayed in Figure 3A. For both the CRISPR-RPA and CRISPR-LAMP onepot visual detections, after the template DNA was added and the tubes were incubated at a suitable temperature in the reaction order, the rapid reaction of pUC57-p72 DNA occurred and a strong fluorescence signal appeared while other genomic DNA or cDNA showed no signal (Figures 3B,C), which revealed there was no cross-reaction with other viruses. The results  (5, 10, 15, 20, 25, and 30 min). The fluorescence signal monochrome and pseudo green images were taken from Tanon 2000M camera, and the images at the bottom were taken behind the UV/blue light resistant observation window. Analysis of the value for the fluorescence image using the ImageJ software. Each measuring was run with three replicates (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant. Error bars represent the means ± s.d. from replicates. The unpaired two-tailed t-test was used to analyze the statistical significance.
demonstrated that both CRISPR-RPA and CRISPR-LAMP onepot visual detection systems could be used for specific detection of ASFV.
While estimating the sensitivity and specificity of the one-pot visual detection system, it showed high specificity and accuracy, as well as low detection limit, particularly the limit of CRISPR-LAMP was much lower.
The limits of the LAMP-CRISPR detection system were determined using a 10-fold serial diluted pUC57-p72 DNA template, ranging from 7 × 10 9 to 7 × 10 0 copies/µl. The results showed that the developed LAMP-CRISPR one-pot visual detection system can detect as low as 7 × 10 2 copies/µl of the dsDNA template ( Figure 4C). Using diluted ASFV B646L gene plasmid reference material as a template, the limit of one-pot visual detection system based on the LAMP-CRISPR was 5.8 × 10 2 copies/µl within 60 min, as shown in (Figure 4D).

Evaluating Consistency Between the One-Pot Visual Detection and qPCR
We further compared one-pot visual detection with the quantitative PCR (qPCR), the most commonly used detection method, as the gold standard. The same target on p72 gene was used for the RPA and LAMP amplification to determine the analytical sensitivity of the qPCR. Serial dilutions were prepared from 10 9 to 10 0 copies/µl. The result showed that the limit of qPCR detection was 7×10 2 copies/µl (Figure 3D). Compared to qPCR detection, the limit of one-pot detection was similar and the CRISPR-LAMP could attain a sensitivity of 5.8 × 10 2 copies/µl. Moreover, both the CRISP-RPA and CRISPR-LAMP detection systems spent lesser detection time than the qPCR.

DISCUSSION
Since 2018, the rapid outbreak of ASF in China and in a number of other countries has resulted in tremendous economic losses (D) qPCR for 10-fold serial diluted pUC57-p72 DNA (qPCR was processed using GraphPad 8.0). Analysis of the value for the fluorescence image used the ImageJ software. Each measuring was run with three replicates (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant. Error bars represent the means ± s.d. from replicates. The unpaired two-tailed t-test was used to analyze the statistical significance. (51). At present, molecular diagnostic techniques for detecting ASFV mainly rely on two OIE-recommended conventional and real-time qPCR technique methods. Although these techniques have been widely validated and are useful tools for detecting this disease, they remain inconvenient because of expensive instruments and professional operation systems. The CRISPR/Cas systems are revolutionary tools allowing for precise genome engineering, transcription regulation, and many other applications (52,53). The Cas12a recognizes specific dsDNA sequences and then non-specifically cleaves the ssDNA, making it particularly suitable for detecting dsDNA viruses.
The cleavage site specificity of the Cas/crRNA complex is determined by the length of the crRNA and the sequence, number, location, and distribution of mismatch (54,55). The 20nt or shorter nucleotide hybridizes with the dsDNA close to PAM region, which determines that CRISPR/Cas has high specificity (56). However, SNV in genome may cause off-target and false negatives when mutating on the target or the PAM (54,57). The introduction of PAM through pre-amplification and multi-crRNA strategies may break the limitation caused by both PAM dependence and off-target, helping to improve the specificity of the detection system. The Cas12a enzyme itself has weak collateral cleavage activity; thus, it can only achieve low detection sensitivity without pre-amplification (27,58). Therefore, in combination with isothermal amplification, such as RPA, LAMP, strand displacement amplification (SDA) (59)(60)(61), rolling circle amplification (RCA) (62,63), exponential amplification reaction (EXPAR) (64,65), and recombinase-aided amplification (RAA) (48,(66)(67)(68), we found that Cas12a was able to detect pathogens with high sensitivity and precision. In this study, we found that the CRISPR-RPA system is not highly sensitive, but it is faster than CRISPR-LAMP, which means that CRISPR-RPA is possible for rapid qualitative detection on the grassroots level.
Aerosol pollution, improper operation, and complex detection environment may unavoidably cause false positives in point-of-care testing (POCT), and two separate steps in CRISPRbased detection may make it more serious. However, one-pot detection could avoid false positives and high background positives caused by cross contamination (69)(70)(71)(72). This means that template amplification, Cas-mediated enzyme cleavage reaction, and signal output are completed in one tube without opening or closing the cover (71,73,74). Moreover, to report the presence of target DNA, ssDNA probe linking a fluorophore (FAM in this study) and a quencher was used. After the The sensitivity of CRISPR-LAMP one-pot detection with 10-fold serial diluted ASFV B646L gene plasmid reference material. Analysis of the value for the fluorescence image using the ImageJ software. Each measuring was run with three replicates (n = 3). *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant. Error bars represent the means ± s.d. from replicates. The unpaired two-tailed t-test was used to analyze the statistical significance.
indiscriminate cleavage was triggered, the fluorophore on the ssDNA probe was released and detected by specific wavelength light, making it visible and suitable for POCT, thereby avoiding the problems described previously (74)(75)(76). The contradiction between sensitivity and specificity in the detection process is thus solved to a certain extent.
Depending on expensive thermocyclers and skilled operator, PCR as the gold standard is difficult to popularize in POCT. Therefore, lower cost and personnel requirement detection methods could be more advantageous and required. Stripbased lateral flow assay (LFA) is an option with low cost (20,77,78), but opening tube while inserting the strip makes it less suitable. Therefore, the CRISPR-isothermal amplification method used in this study shows more advantages in POCT. A hand warmer and water bath, which gets rid of thermocyclers, can satisfy the meet for pre-amplification. Besides, the fixed wavelength illuminant makes the results more intuitive, clear and reduces the dependency on skilled operators, which greatly reduces the cost of POCT.
In conclusion, a one-pot visual detection system was established and used for the rapid, sensitive, specific, and low-cost detection of ASFV. This integration has great potential for POCT detection of ASFV and other DNAbased pathogens, which could be an effective way for the timely monitoring of ASFV to prevent its occurrence and spread at an early stage.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding authors.

AUTHOR CONTRIBUTIONS
KX and MZ contributed to the conceptualization and biosensor design, fabrication, and manuscript preparation. CQ performed the experiments, data curation, and drafted the manuscript. JL, WZ, JD, HZ, JZ, YK, LL, GS, ZL, HL, SJ, and KC performed formal analysis, validation support, and manuscript proofreading. XD and HM reviewed, edited, and supervised this work. All authors have read and agreed to the published version of the manuscript.