Exploring the Trans-Cleavage Activity with Rolling Circle Amplification for Fast Detection of miRNA

MicroRNAs (miRNAs) are a class of endogenous short noncoding RNA. They regulate gene expression and function, essential to biological processes. It is necessary to develop an efficient detection method to determine these valuable biomarkers for the diagnosis of cancers. In this paper, we proposed a general and rapid method for sensitive and quantitative detection of miRNA by combining CRISPR–Cas12a and rolling circle amplification (RCA) with the precircularized probe. Eventually, the detection of miRNA-21 could be completed in 70 min with a limit of detection of 8.1 pM with high specificity. The reaction time was reduced by almost 4 h from more than 5 h to 70 min, which makes detection more efficient. This design improves the efficiency of CRISPR–Cas and RCA-based sensing strategy and shows great potential in lab-based detection and point-of-care test.


Introduction
MicroRNAs (miRNAs) are a class of endogenous short non coding RNA.They regulate gene expression and function, essen tial to biological processes [1].Aberrant miRNA expression has been reported in all types of tumorassociated diseases [2], in dicating that miRNAs have the potential to be valuable biomark ers for the diagnosis of cancers.However, due to their sequence similarity, small size, and degradability, rapid and sensitive quan tification of miRNAs is difficult.In recent years, clustered reg ularly interspaced short palindromic repeats (CRISPR) and CRISPR associated proteins (Cas) have been widely used as ef ficient tools in nucleic acidrelated detection, especially in the determination of viral infections [3].Several nucleic acid am plification methods were generally combined with the CRISPR-Cas system to obtain higher sensitivity, such as polymerase chain reaction [4], recombinase polymerase amplification [5] and loop mediated isothermal amplification [6,7].In these meth ods, at least 1 pair of primers were designed for target detection.However, miRNAs are short singlestranded RNAs with a length of 18 to 25 nt, disallowing these methods to be directly applied for their detection.
To overcome this limitation, several miRNA detection meth ods based on short nucleic acid amplification techniques have been developed.For example, a method based on rolling circle transcriptionCas12a was reported by Wang and coworkers [8].A method based on cascade amplification with a DNAzyme, an RNA ligase, an RNA polymerase, and Cas12a was reported by Sun and coworkers [9].However, the amplification products of these 2 methods are RNAs, which are less stable and more susceptible to hydrolysis than DNA.In addition, several nucleic acid amplification methods with doublestranded DNA (dsDNA) products as an activator chain for Cas12a were reported [10][11][12].Nevertheless, the protospacer adjacent motif (PAM) sequence requires to be considered when dsDNA hybridizes to CRISPR RNA-Cas12a (crRNA-Cas12a) complex, leading to complex oligo nucleotide design.Although the targeting range of CRISPR-Cas12a can be improved by adapting to alternative the PAM se quences, the cleavage efficiency rate is lowered [13].As a com parison, a singlestranded DNA (ssDNA) activator can hybridize to a crRNA-Cas12a complex even without a PAM sequence.
Rolling circle amplification (RCA) is generally recognized as an efficient isothermal amplification technique, amplifying a short DNA or RNA to form massive ssDNA using a DNA pol ymerase [14,15].It is generally used for the detection of short singlestranded nucleic acids such as miRNAs [16][17][18][19][20].Although it is a robust amplification method, its application is limited due to its detection time usually exceeding 5 h [8,[21][22][23].
As an effective nuclease, the CRISPR-Cas system is known to recognize and hybridize to target nucleic acids with high specificity.Unlike Cas13a, Cas12a's target and probe are both DNA, which increases the detection assay's robustness and low ers the cost greatly.Therefore, using Cas12a to detect miRNA would be more trustworthy.In the case of the product of iso thermal amplification is designed as the activator of Cas12a trans cleavage activity, nonspecific amplification product cannot active the Cas12a fluorescence reporter system due to the specifically hybridization of Cas12a-crRNA complex to trigger the chain.In addition, Cas12a is an extremely efficient nuclease, ensuring a short reaction time.Therefore, combining RCA isothermal amplification techniques with CRISPR-Cas system would pro mote the performance of sensitivity, specificity, and reaction time and make it possible to develop highly sensitive, specific, and efficient biosensors for the detection of miRNAs.Recently, several CRISPRassisted strategies for the detection of miRNA have been proposed based on rolling circle transcription and RCA [8,24,25].Even though a favorable detection performance is obtained, the detection time (several hours) of those strategies is too long to meet the need for pointofcare testing, which is precisely the most widely used field of isothermal amplification.
In this work, we aimed to develop a general and rapid method for sensitive quantitative detection of shortstranded nucleic acids.We combined the CRISPR-Cas12a fluorescence reporting system with RCA isothermal amplification with a precircularized probe (Fig. 1) and named it rapid RCA-Cas reaction (rRCA-Cas).As a result, the detection of miRNA could be achieved by monitoring the fluorescence signal with the limit of detection (LOD) of 8.1 pM.In addition, the reaction time was reduced by almost 4 h from more than 5 h to 70 min.
Fluorescence sensitivity assay of Cas12a fluorescence reporter system compared with SYBR Green dye assay and fluorophore-labeled assay Cas12a fluorescence reporter system: Synthetic trigger chain was diluted into different concentrations from 0 to 1 μΜ.After a 10min preincubation of Cas12a and crRNA at room tem perature, the 100μl transcleavage reaction started by mixing all components, including 1× NEB CutSmart™ buffer (50 mM KAc, 20 mM trisHCl, 10 mM MgCl 2 , and 100 μg/ml Bovine serum albumin (BSA) [pH 7.9 at 25 °C]), 30 nM LbCas12a-crRNA complex, 200 nM FQ probe, and various concentrations of trigger chain.The fluorescence intensity of the reaction was monitored by a fluorescence plate reader for 15 min at 37 °C.
SYBR Green dye assay: The 1× SGI dye was added to a total of 100 μl of reaction with different concentrations of synthetic trigger chain from 0 to 1 μΜ.The dye binds to DNA very rapidly, reaching equilibrium within 10 min, leading to a stable fluores cence intensity.After a 10min incubation, the fluorescence intensity was measured.
Fluorophorelabeled assay: Synthetic FAMlabeled trigger chain was diluted into different concentrations from 0 to 1 μΜ.The 100μl solution was seated for 10 min at room temperature before the fluorescence intensity was measured.
The excitation wavelength and emission wavelength of all the fluorescent assays above were set at 492 and 518 nm, respectively.mixed and incubated for 1 h at 60 °C.After heating at 80 °C for 10 min to inactivate the enzyme, the precircularized circular probe was stored at 4 °C until use.
Then, the 10μl amplification product was incubated with a Cas12a fluorescence reporter system consisting of 30 nM Cas12a-crRNA complex, 1× NEB CutSmart™ buffer, and 200 nM FQ probe.The mixture was incubated at 37 °C for 10 min to activate the transcleavage activity of Cas12a.Then, the fluorescence intensity was measured at 492nm excitation wavelength and 518nm emission wavelength.

Data processing
In the experiments for optimization of the reaction parameters, F/F 0 represents the signaltonoise ratio, where F and F 0 are the fluorescence intensity in the presence and absence of miRNA target, respectively.

Results and Discussion
Advantage of Cas12a fluorescence reporter system compared with SYBR Green dye assay and fluorophore-labeled assay SGI dye and related fluorophores were widely used in the mol ecule diagnosis [26][27][28][29][30].We first estimated the sensitivity of the Cas12a fluorescence reporter system, the SGI dye assay, and the FAM fluorophorelabeled assay with the same seriesdiluted target ssDNA.As shown in Fig. 2, 100 nM target could be detected with 1× SGI and the labeled FAM fluorophore; in addi tion, 1 nM target could be detected with the Cas12a fluorescence reporter system.Thus, the Cas12a fluorescence reporter system greatly improved the detection sensitivity for ssDNA target detection.The results indicated the Cas12a fluorescence reporter system has the potential to be used to develop highly sensitive biosensors.In addition, the fluorescence signal from the Cas12a fluorescence reporter system did not increase monotonically with the increase of target concentration.The activity of Cas12a was inhibited at a high concentration of DNA target, which was consistent with our previous report [31].

Schematic and feasibility of rRCA-Cas reaction
In the traditional RCA approach, the padlock probe was de signed to hybridize with miRNA target at both the 3′ and 5′ ends [21,[32][33][34].The following steps including ligation to form a circular DNA template and hydrolyze excess free DNA are time consuming leading to a total detection time of about 5 h [21,22].Herein, we employed a precircularization step to save detection time (Fig. 3A).The padlock probe was designed to hybridize with a helper DNA chain and ligated by T4 DNA ligase.EXO I is often used to hydrolyze ssDNA, and EXO III is used to hydro lyze nonspecific complexes [35][36][37].After hydrolysis of excess free DNA by EXO I and EXO III, a precircularized circular probe acting as a rolling template was obtained.In addition, the acti vated Cas12a can cleavage very fast as an efficient nuclease.Com bining these 2 advantages, this proposed approach can decrease the detection time to about 70 min.
The schematic of rRCA-Cas strategy is shown in Fig. 1.The circular probe DNA was precircularized to shorten the detection time.In the presence of target miRNA (dark orange line), after hybridizing to the precircularization circular probe functional as the template, miRNA was extended by phi29 DNA polymer ase to initiate the RCA.A long ssDNA sequence was generated, containing numerous "trigger" (dark purple line) repeats.Then, the repeat trigger sequences hybridized to the Cas12a-crRNA by base complementation.Then, Cas12a showed transcleavage activity, cleaving the ssDNA around it, including the FQ probe.As a result, the quencher was separated from the fluorophore, increasing the fluorescence signal.
The miRNA21 was used to demonstrate the feasibility of the rRCA-Cas reaction for miRNA detection, and the feasibility was first verified via 12% native polyacrylamide gel electropho resis (PAGE) (Fig. 3B).The helper-padlock complex was formed and showed a band higher and a single padlock chain in lane 4. Ligation of padlock did not change the molecular weight of the complex as well as the band in lane 5. EXO I enzyme catalyzes and removes nucleotides in the 3′ to 5′ direction from linear ssDNA, and EXO III catalyzes and removes nucleotides in the 3′ to 5′ direction from linear or nicked dsDNA.After hydrolysis of EXO I and EXO III, a circular probe was obtained in lane 7.For the miRNA21 detection, 10 nM miRNA target and phi29 DNA polymerase were added, and a long ssDNA product was observed in lane 8.Then, the Cas12a fluorescence reporter sys tem was added, and the long ssDNA amplification product was cleaved by Cas12a, which led to a few kinds of fragments with different molecular weights in lane 9.Then, we employed the fluorescence spectrum to verify the feasibility (Fig. 3C).The fluorescence signal was obtained in the presence of miRNA tar get (blue trace).However, in the absence of miRNA target, the RCA was not initiated, resulting in no generation of the long ssDNA.The Cas12a was inactive and no fluorescence signal increased (red trace).In addition, another circularization method was investigated in which the padlock was directly circularized by the CircLigase enzyme without the helper chain (Fig. S1).It proved to be as effective as the T4 DNA ligasebased method described above.In consideration of cost, we chose T4 DNA ligasebased method for further study.

Optimization of rRCA-Cas reaction
We further optimized the reaction system to achieve the best sensitivity performance.Firstly, the impact of EXO I and EXO III was investigated (Fig. 4A).EXO enzymes were essential to decrease background signal intensity by hydrolyzing free DNAs and the helper chain on the circular probe.Secondly, we esti mated the effect of circular probe concentration (Fig. 4B).The results showed that the fluorescence intensity increased as the concentration of the circular probe increased, and 50 nM cir cular probe was selected for further experiment.Then, the reac tion time of RCA was evaluated, and the results indicated that 1 h was optimal by F/F 0 value (Fig. 4C).The concentration of Cas12a-crRNA complex in the Cas12a fluorescence reporter system was also evaluated (Fig. 4D).The fluorescence intensity increased as the concentration of the Cas12a-crRNA increased, and 240 nM Cas12a-crRNA complex had the best F/F 0 value.However, 30 nM complex was used for further experiment for cost consideration.Moreover, we used Bst 2.0 DNA polymerase to replace phi29 DNA polymerase, so that the RCA reaction could be carried out at 65 °C instead of 37 °C (Fig. 4E).However, the amplification efficiency of Bst 2.0 DNA polymerase was lower than that of phi29 DNA polymerase in our assay.In addi tion, RCA at 37 °C is consistent with the reaction temperature of Cas12a.Thus, phi29 DNA polymerase was used for further study.We further investigate the impact of the storage time on the stability of the circular probe (Fig. 4F).The fluorescence intensity kept stable within 5 d.After 20 d of storage at 4 °C, there is a modest drop in fluorescence intensity, indicating that the circular probe has degraded slightly.Under the optimal con ditions determined above, the whole assay was performed at 37 °C and the detection time was 70 min (1 h for RCA and 10 min for the Cas12a fluorescence reporter system).

Sensitivity, specificity, and real-sample test of rRCA-Cas reaction
Different concentrations of synthesized target miRNAs (100 nM, 10 nM, 5 nM, 1 nM, 500 pM, 100 pM, 10 pM, and 1 pM) were used to evaluate the sensitivity (Fig. 5A).The fluorescence inten sity was measured for 15 min for quantitative analysis following the addition of the Cas12a fluorescent system.Fluorescence intensity was plotted against miRNA21 concentrations, and the regression of the best linear fit was between 100 pM and 10 nM (Fig. 5B).Based on a 3*SD/slope, the LOD was determined to be 8.1 pM.Although this method could not achieve a higher sensitivity than some reported nanomaterialsbased and elec trochemistrybased ones for the detection of miRNAs [38][39][40], the LOD as low as 8.1 pM of this method is comparable to fluo rescence detection comparing to previous work [8,22,24,[41][42][43][44]. Most importantly, our detection time is less than the vast major ity of reported biosensors based on RCA (see Table S2 for details).
To investigate the specificity of the rRCA-Cas reaction, a variety of miRNAs including miRNA21, miRNA126, miRNA 122, and miRNA155 were tested, which have the potential to be the biomarkers of colorectal cancer, venous thromboembo lism of persistent, liver diseases, and Bcell malignancies.The concentrations of these miRNAs were set to be the same at 5 nM to compare the fluorescence signal intensity.As seen in Fig. 5C, miRNA21 had a fluorescence signal that was 19.3fold (P < 0.0001) than miRNA126, miRNA122, and miRNA155, showing this method successfully achieved high specificity for miRNA21.Given the high specificity of the DNA ligase, the high specificity exhibited by the reaction was expected [45,46].
The fluorescence signal of miRNA21 was 19.3fold higher (P < 0.0001), while no fluorescence signals were detected for miRNA126, miRNA122, and miRNA155, indicating that high selectivity of miRNA21 was achieved for this strategy.Both of the 2 padlock precircularization methods-T4 DNA ligasebased and CircLigasebased-showed high specificity (Fig. S1D).
To verify the feasibility and investigate the matrix effect of rRCA-Cas reaction in real samples, we detected different con centration of miRNAs in 1% human serum.The workflow is simple and the miRNA can be detected in 70 min (Fig. 5D).Fluorescence intensity was plotted against miRNA21 concen trations, and the regression of the best linear fit was between 200 pM and 10 nM (Fig. 5, E and F).The recovery of the strat egy was investigated by spiking various concentrations of miRNA21 in 1% human serum.The estimated recoveries ranged from 89.5% to 102.3% (Table S3).

Conclusion
In this study, we proposed a detection strategy for achieving rapid determination of miRNA via the CRISPR-Cas12a fluo rescence reporting system coupled with a precircularized probe in RCA.Finally, with a LOD of 8.1 pM and high specificity, miRNA21 was quantitatively identified in 70 min at 37 °C.The reaction time was reduced by almost 4 h from more than 5 h to 70 min by precircularizing the padlock to obtain the circular probe, which makes detection more efficient.Moreover, we fully illustrated the advantages of using Cas12a as an additional module for isothermal amplification, besides the improved sensitivity compared to the SYBR Green and fluorophore.In addition, the trigger sequence can be designed as the same sequence in our proposed approach, and the identification of different miRNAs can be achieved by simply replacing the pad lock chain.The Cas12a fluorescence reporting system could be used as a general additional module without any change, which made our approach more convenient and timesaving.
In addition, we have reported an aptamerbased method for the detection of βestrodial [47], which was based on the aptamer structureswitch binding to βestrodial.It can serve as a recog nition element for nonnucleic acid molecules to develop the biosensor.Furthermore, the aptamerbased structureswitch recognition element could be integrated with the proposed rRCA-Cas reaction, developing of nonnucleic acid molecule biosensors besides the detection of miRNAs.

Fig. 1 .
Fig.1.Schematic of the rRCA-Cas reaction.After hybridizing to the precircularized circular probe, the target miRNA is amplified into a long ssDNA product by RCA with massive trigger sequences repeats.The trigger sequence is designed as the activator of Cas12a for showing trans-cleavage activity.Then, the trigger chain is recognized by Cas12a-crRNA duplex and hybridizes to it.Cas12a was then triggered to demonstrate trans-cleavage activity, F-Q probe (an ssDNA probe labeled with a fluorophore at the 5′ end and a quencher at the 3′ end) was cleaved, resulting in separation of the quencher from the fluorophore, generating the fluorescence signal.

Fig. 2 .
Fig. 2. Estimation of the sensitivity of Cas12a fluorescence reporter system, SYBR Green, and FAM fluorophore with the same diluted target ssDNA.Sensitivity with Cas12a fluorescence reporter system, 1× SGI, and labeled fluorophore were tested.Error bars present the SD with 3 individual tests.NC, background signal with distilled water as a negative control.****: P < 0.0001.ns, no significant difference.

Fig. 3 .
Fig. 3. Feasibility of rRCA-Cas reaction.(A) The workflow of precircularization of the padlock to obtain circular probe.(B) Verification by the 12% native PAGE image.Lane 1: ssDNA ladder.Lane 2: 1 μM helper chain.Lane 3: 1 μM linear padlock chain.Lane 4: After incubation of helper and padlock chain.Lane 5: Padlock was circularized to circular probe by T4 DNA ligase with helper hybridized on it.Lanes 6 and 7: After EXO I and EXO III hydrolysis, circular probe without helper on it was observed.Lane 8: The RCA product of 10 nM target miRNA-21.Lane 9: The RCA product was cleaved by Cas12a/crRNA complex.(C) Verification by the fluorescence spectrum.The blue trace represents the signal of the entire rRCA-Cas reaction in the presence of the target, which is the same as lane 9 in the PAGE image.The orange trace represents the signal in the absence of the target.NC indicates the blank signal without the target in the reaction.

Fig. 4 .
Fig. 4. Optimization of rRCA-Cas reaction.(A) Evaluation of the impact of EXO I and III on rRCA-Cas reaction.(B) The fluorescence intensity and F/F 0 values of various circular probe concentrations.(C) The fluorescence intensity and F/F 0 values of various amplification times of RCA reaction.(D) The fluorescence intensity and F/F 0 values of various Cas12a-crRNA complex concentrations.(E) The performances with different DNA polymerases used in the rRCA-Cas reaction.(F) The fluorescence intensity and F/F 0 values of various storage times of circular probe.NC indicates the blank signal without the target in the reaction.F/F 0 means the fluorescence intensity ratio of the group with miRNA target and the group without miRNA target (NC), which represents the signal-to-noise ratio.Error bars present the SD with 3 individual tests.

Fig. 5 .
Fig. 5.The performance of the rRCA-Cas reaction for detecting miRNA-21.(A) Fluorescence intensity at varied target miRNA concentrations ranging from 1 pM to 100 nM under buffer conditions.(B) The calibration curve of fluorescence intensity and varied target miRNA concentrations.Inset: Linear range from 100 pM to 10 nM, R 2 = 0.9888.(C) Specificity evaluation for 5 nM miRNAs detection with different miRNA species (miRNA-21, miRNA-126, miRNA-155, and miRNA-122).****: P < 0.0001.(D) Workflow for miRNA detection in serum.(E) Fluorescence intensity at varied target miRNA concentrations ranging from100 pM to 10 nM in 1% human serum.(F) The linear range from 200 pM to 10 nM regressed from the data in (E), R 2 = 0.9775.Error bars present the SD with 3 individual tests.NC indicates the blank signal without the target in the reaction.