The sources of MCMV and SCMV were kept in our laboratory. Both viruses were propagated on maize inbred line B73 and cv. Nongda 108 plants, which were grown in a growth chamber (28 °C 16h light and 22°C 8h night cycles) for virus propagation. The first true leaves of one-week-old maize seedlings were rub-inoculated with the homogenized MCMV- and SCMV- infected maize leaf tissues in 0.01 M phosphate buffer [0.01 M KH₂PO₄: 0.01 M Na₂HPO_{4 =} 49: 51 (v/v), pH 7.0] at a ratio of 1:10 (g/mL). The systemically infected leaves were harvested at about 10 days post inoculation (dpi). The total RNA of cucumber mosaic virus (CMV), tobacco mosaic virus (TMV), tomato ringspot virus virus (TRSV), tomato black ring virus (TBRV) were stored in our laboratory.

Extraction of total RNA from maize leaf tissue (~100 mg) was performed using EasyPure ^® RNA Purification Kit (TransGen, Beijing, China) according to the manufacturer’s instructions. The extracted nucleic acid was dissolved in 50 μL nuclease-free water and stored at -80°C prior to testing. Total RNA was quantified using a BioTek Epoch spectrophotometer (BioTek, Vermont, USA).

The crude extracts of total RNA from plant materials was extracted with modified alkaline polyethylene glycol (PEG) extraction method (Chomczynski and Rymaszewski, 2006; Huang et al., 2013; Silva et al., 2018). In brief, the fresh leaf samples (about ϕ 8 mm size) were punched with the lids of 1.5 mL tubes, crushed with quartz sands using a plastic pestle, immersed in 100 μL of freshly prepared alkaline-PEG buffer [6% PEG 200 (Solarbio, Beijing, China) with 20 mM NaOH] (Huang et al., 2013; Silva et al., 2018), and incubated at room temperature for 4 min. The plant extract supernatants were tested immediately or kept on ice until further use.

RT-RAA primers and real time fluorescent probe were designed from the conserved coat protein region of MCMV following multiple sequence alignment of the available virus sequences in the GenBank database ( Figure 1 ). The primers and probe were BLASTed against the GenBank to exclude other plant viruses, including SCMV, cucumber mosaic virus (CMV), tobacco mosaic virus (TMV), tomato ringspot virus (TRSV), tomato black ring virus (TBRV). The forward primer MCMV-F (30 nt) and reverse primer (33 nt) ( Table 1 ) produced target amplicons of 116 bp. The real-time fluorescent probe consists of an oligonucleotide with homology to the target amplicon that contains a dSpacer which replaces a nucleotide in the target sequence flanked by a dT-FAM (fluorophore) and corresponding dT-BHQ1 (quencher). In addition, C3-spacer as a blocker was labelled at the 3’-end to prevent polymerase extension from the 3’-terminus. When the target amplicon was produced, this fluorescent probe is cleaved by the E. coli exonuclease III at the abasic site to separate the fluorophore “FAM” from the quencher “BHQ1” and generate an extensible 3’-OH group for polymerization, thus generating fluorescent signals. All primers and probe were synthesized by Sangon Biotech (Shanghai, China).

The crRNA was designed to recognize a site specific to MCMV genome and to be homologous to a region within the amplicon between the RT-RAA primer pair binding sequences. The Cas12a/crRNAs recognize a 20-bp target sequence adjacent to a PAM site (TTTN or NAAA) site and were designed from the highly conserved region of each virus as found in the GenBank (NCBI). There was one PAM site (GAAA) in the RT-RAA amplicons of MCMV, whose reverse complement sequence is TTTC, so the corresponding reverse and complementary sequence was reversely transcribed as the DNA template for crRNA. The DNA template for crRNA was further checked its specificity by BLAST against the GenBank. The DNA sequences and modifications are outlined in Table 1 . Fluorescent quencher reporter (FQ reporter: FAM-TTATT-BHQ1), lateral flow assay reporter (LF reporter: FAM-T₂₀-biotin) and crRNA were synthesized by General BioL. (Anhui, China).

The real-time fluorescent detection of the viral RNA was performed with fluorescent RT-RAA kits according to the instructions of the RT-RAA fluorescent kit (Cat No. F00R01, Jiangsu Qitian Gene Biotechnology Co., Ltd., Jiangsu, China). Briefly, a master mix containing 25.0 μL of supplied rehydration buffer, 10.0 μL of DNase/RNase-free deionized water (Cat. No. RT121, Tiangen, Beijing, China), 2.4 μL of forward primer (10 μM), 2.4 μL of reverse primer (10 μM), and 2.5 μL of magnesium acetate (280 mM) was added to a RT-RAA pellet to dissolved the enzyme. And 2 μL of extracted RNA was added and mixed. Next, 2.5 μL of magnesium acetate (280 mM) was added on the tube lid, which was mixed with the RT-RAA reaction buffer with centrifugation. The Eppendorf tubes were put in a LightCycler 480 (Roche, USA) to record the emitted fluorescence signals. The excited and emitted wavelength are 488/520 nm, respectively. RNA extracted from healthy maize leaves and DNase/RNase-free deionized water were included as the negative control alongside tested samples.

The pre-amplification of the viral RNA for CRISPR/Cas12a detection was performed with RT-RAA kits according to the instructions of the RT-RAA basic kit (Cat. No. B00R00, Jiangsu Qitian Gene Biotechnology Co., Ltd., Jiangsu, China) with modifications. Briefly, a master mix containing 25.0 μL of supplied rehydration buffer, 10.0 μL of DNase/RNase-free deionized water (Cat. No. RT121, Tiangen, Beijing, China), 2.4 μL of forward primer (10 μM), 2.4 μL of reverse primer (10 μM), and 2.5 uL of magnesium acetate (280 mM) was added to a RT-RAA pellet to dissolved the enzyme, A reconstituted RT-RAA reaction aliquot of 9.0 μL was transferred to new PCR tubes, then 2 μL of extracted RNA was added to each reaction, and mixed by carefully pipetting up and down. The reactions were incubated at 39°C on a self-developed thermal block for 15-30 min. RNA extracted from healthy maize leaves and DNase/RNase-free deionized water were included as the negative control alongside tested samples.

The Cas12a-mediated fluorescent detection contained 1×NEBuffer 3.1 (Cat. No. B7203, New England Biolabs, Ipswich, MA, USA), 0.1 μM of EnGenLba Cas12a (Cat. No. M0653T, New England Biolabs, Ipswich, MA, USA), 0.12 μM of crRNA, 0.1 μM FQ reporter, 8U of RNase inhibitor (Cat. No. NG209, Tiangen, Beijing, China), 2.0 mM DTT (Cat. No. 43816, Sigma Aldrich, St. Louis, MO, USA) and 1 μL RT-RAA amplicons in 20 μL reaction volume. The reaction was performed at 37 °C in a LightCycler 480 (Roche, USA) to record the emitted fluorescence signals. The excited and emitted wavelength are 488/520 nm, respectively.

The Cas12a-mediated lateral flow assay contained 1×NEBuffer3.1 (New England Biolabs, Ipswich, MA, USA), 0.1 μM of EnGenLba Cas12a (New England Biolabs, Ipswich, MA, USA), 0.12 μM of crRNA, 1.0 μM LF reporter, 8U of RNase inhibitor (Tiangen, Beijing, China), 2.0 mM DTT (Sigma Aldrich, St. Louis, MO, USA) and 1 μL RT-RAA amplicons in 20 μL reaction volume. The reaction was performed at 37 °C in an incubator for 30 min and the products were detected with lateral flow strips (Cat. No. JY0301, Tiosbio Biotechnology Co, Ltd., Beijing, China).

A master mix containing 25.0 μL of supplied rehydration buffer, 6.0 μL of nuclease free water, 4.0 μL LF reporter (10 μM), 2.4 μL of forward primer (10 μM), 2.4 μL of reverse primer (10 μM) was added to an RT-RAA pellet to dissolve the enzyme, followed by the addition of 2.5 uL of magnesium acetate (280 mM), then 9 μL of the reconstituted RT-RAA reaction buffer was distributed to the bottom of a new PCR tube. Meanwhile, the Cas12a reaction buffer containing 7.5 μL NEBuffer 3.1 (10×), 2.5 μL EnGenLba Cas12a (5 μM), 2.5 μL DTT (0.1 M), 2.5 μL RNase inhibitor (40U) and 5.0 μL crRNA (5 μM) was distributed into 4 μL aliquots, which was added on the PCR tube lid. If a larger number of reactions are needed, the master mix volume and pellets are scaled up accordingly. One microliter RNA sample was added into the bottom of the PCR tube, mixed by carefully pipetting up and down, then the PCR tubes were gently closed and put on an incubator at 37°C for 20 min. Subsequently, the CRISPR reagents pre-placed on the PCR tube lid were mixed with the RPA reaction buffer by inverting and centrifuging the tube, and incubated at 37°C for another 20 min. When the FQ reporter was used, the PCR tubes were put in a portable fluorimeter or real-time PCR instrument. When the LF reporter was used, the PCR tubes were put in an incubator and lateral flow strip was directly inserted into the reaction buffer after 85 μL water was added.

Ten-fold serial dilutions of total RNA (96.0 ng/μL) and plasmid genomic DNA containing the MCMV coat protein gene (2.5×10⁷ copies) were used to evaluate the detection sensitivity of the real-time fluorescent RT-RAA assay and one-tube one-step RT-RAA/CRISPR-Cas12a detection in triplicate. The previously reported primers (MCMV-PF1/MCMV-PR1) and TaqMan probe (MCMV-Probe) ( Table 1 ) of the TaqMan real-time RT-PCR method (Zhang et al., 2016) was employed as a control. The TaqMan real-time RT-PCR amplification was performed in a 20.0 μL of reaction volume containing 10 μL of 2×PerfectStart™ Probe One-Step qPCR SuperMix (Cat. AQ221, TransGen, Beijing, China), 0.4 μL of TransScript ^®II Probe One-step RT/RI Enzyme Mix (Cat. AQ221, TransGen, Beijing, China), 0.4 μL of MCMV-PF1 (10 μM), 0.4 μL of MCMV-PR1 (10 μM), 1 μL of MCMV-probe (1 μM), 6.8 μL of RNase-free and 1.0 μL of RNA, were run on a LightCycler 480 (Roche, USA).

MCMV inoculated maize leaves were harvested at 5, 7 and 12 dpi, and the fresh maize leaves were punched with the lids of 1.5 mL tubes, and the crude extracts of total RNA from plant materials were prepared using freshly prepared alkaline-PEG buffer. One microliter of the supernatant was detected using this one-tube one-step RT-RAA/Cas12a assay. Three biological repeats were tested for each sample at different dpi.

As illustrated in Figure 2 , this Cas12a-based one-tube plant viral RNA detection platform using lateral flow strips integrates (A) RNA extraction, (B) RT-RAA pre-amplification of plant viral RNA, (C) sequence-specific recognition of amplicons and non-specific cleavage of LF reporters by Cas12a/crRNA, and (D) visual detection of cleaved product from LF reporter. Both RT-RAA amplification reaction buffer and Cas12a/crRNA buffer were loaded into the same tube before the reaction started, and the Cas12a/crRNA buffer was mixed with RT-RAA pre-amplification reaction buffer by shaking operation after the RT-RAA finished. When the Cas12a reaction ended, water was added into the tube to dilute the reaction buffer, and a lateral flow strip was inserted to detect the cleaved products of LF reporters. As shown in Figure 2C , the cleavage ability of CRISPR/Cas12 complex is activated when the crRNA specifically complements with the target DNA amplicons, and the LF reporters modified with both FAM- and biotin-group was cleaved to produce molecules with free FAM- or biotin-group. When the lateral flow strip contacts with the diluted Cas12a reaction buffer, the FAM-group in the buffer conjugates with the anti-FITC antibody on the surface of AuNPs immobilized at the conjugation pad. For negative sample, the LF reporter is intact, and the FAM-AuNPs conjugates have biotin-group, which would be captured by the streptavidin immobilized on the control area to form the control band. For positive sample, the FAM-AuNPs conjugate passes through the control area due to the lack of biotin group, arriving at the test area to form positive band ( Figure 2D ). Therefore, for the negative sample, only control band appears, while both test band and control band appear for the positive control, or only test band for the strong positive control. The whole experiment including both RT-RAA and Cas12a reaction can be performed on a home-made portable incubator between 37~42°C. By employing the fast plant RNA extraction procedure, the total time from sample preparation to results readout is about 1 h.

The specificity of RT-RAA primers was first evaluated using the real-time fluorescent RT-RAA assay. As shown in Figure 1 , the forward and reverse primers covered the conserved sequences of all the MCMV strains, while the sequences after the “dSpacer” group have one or two different bases. The real-time fluorescence signals showed that only RNA extracted from MCMV infected maize leaves produced obvious fluorescence intensity, but no intense fluorescence intensity was produced by the total RNA extracted from SCMV- infected maize leaves, healthy maize leaves and other plant virus infected plant leaves, which demonstrated that the RT-RAA primers are specific for MCMV ( Figure 3A ).

The specificity of RT-RAA primers and crRNA for RT-RAA/CRISPR-Cas12a detection were tested with the CRISPR-Cas12a based fluorescent detection and lateral flow assay. The results showed that other plant viruses, such as SCMV, CMV, TMV, TRV, TBRV produced very low fluorescence signals ( Figure 3B ), while MCMV produced strong fluorescence signals, indicating that the primers and crRNA have good specificity for MCMV. The lateral flow strips results demonstrated that only MCMV showed up on the test band, which was consistent with that of fluorescent detection ( Figure 3C ).

In this study, we tested 5 nt and 20 nt ssDNA modified with FAM- and biotin-group with different concentration to check the effect of LF reporter on the control and test bands. The results showed that 100 nM LF reporter (20 nt) in the final detection buffer is enough to avoid the false positive phenomenon ( Figure 4A ). To optimize the CRISPR/Cas12 reaction time, the lateral flow strips results were compared using MCMV RNA (positive control) and DNase/RNase-free deionized water (negative control) as the template for RT-RAA. The results showed that MCMV RNA as the template could generate obvious test bands with 10 min CRISPR/Cas12a reaction time, while negative control produced no test bands ( Figure 4B ). Although the intensity of test band became darker with the increased CRISPR reaction time, long reaction time may generate non-specific products, and is not favorable for rapid detection. Therefore, 20-30 min is selected as the subsequent Cas12a/crRNA reaction time.

To make this method feasible for the field detection, we adopted a one-tube strategy to avoid the step-by-step (two-step) experimental operation. In this strategy, one-step RT-RAA reagents and CRISPR-Cas12a reagents were put in one tube before the addition of sample RNA, so no lid opening was required during the experiment. To compare the detection efficiency of two-step and one-tube operation, three levels of MCMV RNA were tested using two-step way (10 μL RT-RAA reaction buffer was mixed with 10 μL CRISPR reaction buffer and 2 μL RT-RAA reaction buffer was mixed with 18 μL CRISPR reaction buffer) and one-tube way. There is no big difference amongst the three ways when high amount of MCMV RNA (9.6 ng and 9.6 × 10^-3 ng) was used ( Figure 5 ). When lower MCMV RNA was detected, the one-step operation generated the highest fluorescence signal, while the two-step (2 μL RT-RAA buffer was mixed with 18 μL CRISPR buffer) produced the lowest. Therefore, one-tube operation is a good alternative for the combined one-step RT-RAA with CRISPR-Cas12a detection system.

The results showed that the real-time fluorescent RT-RAA assay could detect the total RNA down to 10^{- 6} dilution of total RNA (9.6× 10^-5ng), whose weak fluorescence signal could be detected ( Figure 6A ). Since the RNA was exponentially amplified by the RAA technique, the onset time of amplification was plotted against the log concentration of the total RNA. The highest RNA concentration (96.0 ng) and the lowest RNA concentration (9.6 × 10^-5 ng) were excluded out of the linear range, and the concentration range from 9.6 ng to 9.6 × 10^-4 ng was linearly fitted with a R² value of 0.963 ( Figure 6B ).

Aa comparison, the TaqMan real-time RT-PCR detection results showed that although the total RNA down to 10^-8 dilution of (9.6 × 10^-7ng) generated fluorescence signal ( Figure 6C ), the Ct value of the 3 replicas is 40, 37.59 and 38.37, respectively. The 10^-7 dilution of total RNA (9.6×10^-6ng) generated fluorescence signal, and the Ct value of the 3 replicas 32.99, 35.29 and 34.94, respectively. The 10^-6 dilution of total RNA (9.6 × 10^-5ng) generated fluorescence signal, and the Ct value of the 3 replicas 30.69, 30.8 and 30.55, respectively. Since Ct value of 35 was regarded as the critical value, the linearity was evaluated using the total RNA from 9.6 ng to 9.6 × 10^-5 ng, and the Ct value was fitted against Log (total RNA) with a R² value of 0.98 ( Figure 6D ).

The sensitivity of one-tube one-step RT-RAA/Cas12a detection for MCMV was evaluated using the ten-fold serially diluted RNA as the template. The results showed that 10^-6 dilution of total RNA (96 fg) can be detected using the fluorescent detection ( Figure 7A ), and 10^-5 dilution of total RNA (0.96 pg) can be obviously detected using the lateral flow assay ( Figure 7B ). Since one-step RT-RAA/Cas12a involved exponential amplification and enzyme/nucleic acid reaction, there is no good mathematic mode to explain the relation between the fluorescence signal and RNA concentration. To determine the absolute sensitivity, plasmid genomic DNA containing MCMV coat protein gene ranging from 2.5 to 2.5×10⁷ DNA copies were tested. Both the fluorescent detection and lateral flow assay could detect as low as 2.5 copies ( Figures 7C, D ).

To check the feasibility of the developed one-tube one-step RT-RAA/Cas12a lateral flow assay, MCMV were tested in inoculated maize leaves at different days post-inoculation (dpi). The crude extracts of total RNA from MCMV inoculated maize leaves produced obvious fluorescence signals, whose intensity increased with longer infection time from 5 to 12 days ( Figure 8A ), while there was no fluorescence signal resulted from the mock inoculated maize leaves ( Figure 8A , Line Mock 1-3). Accordingly, the one-tube one-step RT-RAA/Cas12a based lateral flow assay showed that no test band was detected from the crude extracts of total RNA from mock inoculated healthy maize leaves, but the MCMV inoculated maize leaves at 5 dpi was detected as weak positive ( Figure 8B ). Results of both fluorescent detection and lateral flow assay demonstrated that the amount of MCMV in maize leaves increased with the longer infection time, indicating that this one-tube one-step RT-RAA/CRISPR-Cas12a can be used to detect the field plant sample.