Removing self-cutting in CRISPR tests for faster one-step genetic detection

In this study, human nasal swab specimens for SARS-CoV-2 detection were collected and provided by the College of Medicine and Health Science under a protocol approved by the ethics committee at Wuhan Polytechnic University. Human vaginal swab samples for HPV16 detection were obtained from the pharmacy department at Henan Cancer Hospital, following an approved ethics protocol from Zhengzhou University. Additionally, human oral cancer tissue samples were collected from cancer patients and healthy controls at the First Affiliated Hospital of Zhengzhou University, under another ethics committee-approved protocol. All participants provided written informed consent.

FAM-labeled ssDNA probes, DNA and RNA oligonucleotides, and gene fragments were synthesized by Sangon Biotech (Shanghai, China). The sequences of all the nucleic acids are listed in. Buffer preparation reagents were purchased from Sinopharm Chemical Reagent Co., Ltd. (Beijing, China). NEB Buffer 2.1 was supplied by New England Biolabs (Beijing, China). The RNA extraction miRCURY LNA RT Kit was obtained from Qiagen. RT–RPA and RPA kits were provided by Weifang Amp-Future Biotech (Changzhou, China), and the LAMP kit was sourced from New England Biolabs (Beijing, China). Proteinase K (Takara, Dalian, China) was used for protein digestion. Ultrapure water was used throughout all experiments, and all DNA/RNA samples were dissolved in DEPC-treated water and stored at −20°C. Supplementary Table S1

The gene fragments of AsCas12a, LbCas12a, and CtCas12a were cloned into the 6×His-pET-28a(+) vector and used to transform Escherichia coli BL21 (DE3). When the culture reached an OD₆₀₀ of 0.8 at 37°C, protein expression was induced by adding 0.5 mM isopropyl β-D-1-thiogalactoside, followed by incubation at 18°C for 16 h. Cells were then harvested by centrifugation. The target proteins were purified from the cell lysate using Ni-NTA resin and eluted with a buffer containing 20 mM Tris–HCl, 500 mM NaCl, and 500 mM imidazole. The eluted protein was further purified by gel filtration (Superdex 200 Increase 10/300 GL) in a buffer containing 20 mM Tris–HCl and 200 mM NaCl (pH 7.5). The final storage buffer consisted of 20 mM Tris–HCl (pH 7.5), 250 mM NaCl, and 5% glycerol. Purified proteins were concentrated using a 100 kDa MWCO filter, quantified by Bradford assay, flash-frozen in liquid nitrogen, and stored at −80°C.

Initially, WT Cas12a and SCas12a RNPs were reconstituted in reaction buffer by combining each Cas12a nuclease with its corresponding full-length or split crRNA at a 1:2 molar ratio, followed by a 20-min incubation at room temperature. A target dsDNA plasmid, P-HPV16, was constructed by inserting a 20-bp HPV16 DNA fragment containing a PAM sequence and another 20-bp fragment lacking a PAM sequence into the pcDNA3.1(+) vector, yielding a plasmid of 5430 bp in total length. The cleavage reaction was carried out in a 10 μl mixture containing 1 μl of 10× NEB Buffer 2.1, 100 nM Cas12a, 200 nM crRNA (or 200 nM split RNA), and 10 nM HPV16 plasmid. The reaction was incubated at 37°C for 30 min and then terminated by heating at 90°C for 10 min. Cleaved DNA fragments were analyzed by 2% agarose gel electrophoresis. The gel was then stained with GoldView and visualized under a UV transilluminator for image capture.

First, wild type (WT) Cas12a and SCas12a were reconstituted by incubating each nuclease with its cognate full-length or split crRNA at a 1:2 molar ratio in reaction buffer for 20 min at room temperature. Three HPV16-derived dsDNA substrates were generated by qPCR: one containing a canonical PAM (TTTG), one with a suboptimal PAM (TTTA), and one lacking any PAM (PAM-free). These dsDNAs served as cleavage substrates in subsequent assays. The qPCR reaction mixture was then combined with the Cas12a cleavage system—comprising 1 μl of 10× NEB Buffer 2.1, 100 nM Cas12a RNP, 200 nM crRNA (or split crRNA), and 10 pM dsDNA substrate—and analyzed in real time on a CFX96 Touch system (Bio-Rad, CA, USA). Control reactions contained identical dsDNA concentrations but omitted Cas12a RNP.

The cleavage reactions were carried out in a final volume of 10 μl containing 1 μl of 10× NEB Buffer 2.1, 250 nM Cas12a enzyme, 500 nM crRNA (or 500 nM split RNA), 10 nM dsDNA with varying PAM sequences, and 1000 nM fluorescent probes. The reactions were monitored at 37°C for AsCas12a and LbCas12a, and at 55°C for CtCas12a using a CFX96 Touch real-time PCR system (Bio-Rad, CA, USA).

One lyophilized RPA pellet was resuspended in 29.4 μl of buffer A, 14.1 μl of nuclease-free water, 2 μl of 10 μM RPA forward primer, and 2 μl of 10 μM RPA reverse primer to prepare the RPA reaction mix, following the manufacturer’s instructions (Weifang Amp-Future Biotech). For the RT–RPA reaction, 0.9 μL of RNase H (50 U/μL stock; New England Biolabs, NEB) and 0.45 μL of SuperScript IV reverse transcriptase (Thermo Fisher) were added to the RPA mixture. The reaction was carried out at 37°C.

One-pot reactions were carried out in 20 μl reaction volumes containing 100 nM AsCas12a RNP, 500 nM FQ ssDNA reporter, dsDNA substrate, and RPA or RT–RPA components. The RNP complex, FQ ssDNA reporter (6 μl), and RPA mixture (10 μl) were added to each reaction well, followed by the addition of 2 μl of reaction buffer and dsDNA activator. Reactions were then monitored using a CFX96 Touch Real-Time PCR System (Bio-Rad, CA, USA) at 37°C.

The one-pot reaction (20 μl total) contained 100 nM AsCas12a RNP, 1 pM dsDNA substrate, and RPA components. Ten microliters of pre-assembled RNP–RPA mixture was added to each well, followed by reaction buffer and 2 μl of dsDNA activator. Reactions were incubated at 37°C for 50 min. At designated time points, reactions were heat-inactivated at 95°C and treated with proteinase K; products were then resolved on a 2% TAE agarose gel. Gels were stained with GoldView and imaged under UV transillumination.

The LAMP/SCas12a one-pot testing was conducted in a 25 μl reaction volume with 100 nM CtCas12a RNP, 500 nM FQ ssDNA reporter, 1.4 mM of each dNTP, 8 mM MgSO₄, 50 mM taurine, 0.32 U/μl Bst 2.0 WarmStart DNA polymerase, 1 × primer (0.2 μM F3/B3 primers, 0.4 μM Loop F/Loop B primers, and 1.6 μM FIP/BIP primers). Additionally, 1 × reaction buffer composed of 20 mM Tris–HCl, 10 mM (NH4)₂SO₄, 2 mM MgSO₄, and 0.1% Tween 20 was included, along with 2 μl of substrates. Fluorescence values were monitored using a CFX96 Touch Real-Time PCR System (Bio-Rad, CA, USA) at 55°C for 60 min.

For the detection of specific DNA targets, primers and a qPCR detection kit were obtained from GenePharma (Suzhou, China). Clinical samples (e.g. HPV16) were analyzed according to the manufacturer’s protocol using a CFX96 Touch Real-Time PCR System (Bio-Rad, CA, USA) with SYBR Green fluorescence. Thermal cycling was performed with an initial denaturation at 94°C for 3 min, followed by 40 cycles of denaturation at 94°C for 12 s and annealing/extension at 62°C for 30 s. RNA target detection (e.g. SARS-CoV-2) was carried out using an RT-qPCR detection kit under the same cycling conditions.

The Limit of Detection (LOD) was determined by performing the trans-cleavage assay using varying concentrations of RNA or DNA targets. The LOD value was calculated using the formula 3σ/slope, where σ denotes the standard deviation of three blank measurements. The distinguishing factor (DF) was calculated as: DF = (F_WT − F_WT0)/(F_SNV − F_SNV0), where F_SNV and F_WT are the fluorescence signals generated by the SNV and WT targets, respectively. F_WT0 and F_SNV0 represent the initial fluorescence signals generated by the SNV and WT targets

No statistical method was used to predetermine sample size. All data were included in the analyses without exclusion. The experiments were not randomized, and the investigators were aware of the group allocations during the experiments and outcome assessments.

The conventional Cas12a system recognizes both the canonical PAM (TTTV) and suboptimal PAMs such as NTTV and TTTN [38], and shows significantly enhanced performance in one-pot reactions when suboptimal PAMs are employed. However, the use of suboptimal PAMs still imposes sequence constraints and retains partial cis-cleavage activity, which can interfere with isothermal amplification processes. Therefore, an ideal Cas12a system for one-pot testing should be independent of PAM requirements and free from cis-cleavage activity.

Recently, we developed the SCas12a assay (Fig. 1a), which combines the Cas12a enzyme with a split crRNA composed of scaffold and spacer RNA, enabling direct RNA detection without reverse transcription [36]. In this study, we re-evaluated the SCas12a system’s capability for DNA detection. First, we tested three Cas12a variants, including As, Lb, and Ct Cas12a, and examined their cis-cleavage activity on a plasmid (Fig. 1b) with or without a canonical PAM. The “double-band” pattern corresponds to supercoiled and linear forms of the plasmid DNA, confirming that plasmids lacking a PAM remained uncleaved. The “single-band” pattern represents exclusively linear DNA, indicating complete cleavage of plasmids bearing a TTTG PAM. These observations were further validated by qPCR (Supplementary Fig. S1). In this experiment, dsDNA substrates—identical in crRNA-targeting sequence but differing in their PAM sequences—were incubated with AsCas12a RNP complexes formed by WT or SCas12a. Only SCas12a targeting PAM-free substrates yielded Ct values identical to the no-enzyme control, confirming complete loss of cis-cleavage activity.

Next, using AsCas12a as an example, we compared the trans-cleavage activity of the WT (Fig. 1c) and the SCas12a (Fig. 1d) on dsDNA substrates containing the same 20-bp targeting sequence, but varying in their PAM sequences or the absence thereof. Consistent with previous reports, WT Cas12a showed faster kinetics and stronger fluorescence signals with canonical (TTTG and TTTA) and three suboptimal PAMs [34], whereas PAM-free substrates exhibited the lowest trans-cleavage activity and kinetics. In contrast, SCas12a unexpectedly exhibited substantially higher trans-cleavage activity against PAM-lacking DNA targets, surpassing even that observed with canonical and suboptimal PAMs in the WT Cas12a system (Fig. 1e). Moreover, this behavior was consistent across all tested Cas12a variants (Fig. 1f–g), indicating that the split-crRNA strategy is universally applicable for PAM-independent and cis-cleavage-free DNA detection, laying the foundation for the one-pot CRISPR diagnostics described in the following sections.

To develop a one-pot CRISPR-based detection assay for trace amounts of nucleic acid molecules, we aimed to establish its design principles and determine the optimal reaction conditions. Initially, we investigated the effects of target length and the location of PAM-free sites on the detection efficiency of the SCas12a system. Accordingly, a series of target DNA fragments ranging from 40 to 200 bp, with PAM-free sites positioned at the head, middle, and tail regions (Fig. 2a), were tested. The results demonstrate that the middle location design showed the slowest kinetics and could not be detected when the length exceeded 120 bp (Fig. 2b). In contrast, the designs with PAM-free sites at the head and tail exhibited faster kinetics and stronger fluorescence signals. Notably, the head location showed the best detection efficiency at 120 bp, which is the optimal targeting length for INA. Furthermore, a parallel comparison of the three models was conducted using an identical 120 bp HPV18 DNA target (Fig. 2c), revealing that the head design demonstrated the fastest reaction kinetics and the highest fluorescence intensity. Accordingly, the design incorporating a PAM-lacking 120 bp DNA target was selected for subsequent development of the one-pot assay, unless otherwise stated.

Next, we evaluated the universality of SCas12a assay by testing it against a variety of DNA targets (Fig. 2d). The results demonstrate that all five selected DNA sequences were efficiently detected. Notably, targets with lower GC content, especially at the 5′-end, showed enhanced detection efficiency. These findings offer valuable insights for guiding the design of subsequent one-pot detection experiments. Finally, using the AlphaFold3 Server, we performed predictive structural modeling to analyze the complex formed by Cas12a with the S0 scaffold and a spacer targeting dsDNA without a PAM (Fig. 2f). The crystal structure of Cas12a in complex with crRNA targeting PAM-containing dsDNA was analyzed (Fig. 2e). The predicted structure of the SCas12a/DNA complex reveals a conformation similar to that observed in the WT Cas12a/DNA complex, enabling efficient access of ssDNA probes to the RuvC domain for trans-cleavage. Furthermore, the non-target strand (NTS) without a PAM sequence extends outward, assisting in dsDNA unwinding. However, longer DNA strands exhibit reduced unwinding efficiency (Supplementary Fig. S2), making detection more challenging. Collectively, these findings elucidate the structural basis for PAM-free detection of DNA using the SCas12a system.

In the conventional one-pot CRISPR assay utilizing Cas12a and a canonical PAM (Fig. 3a), CRISPR-mediated detection and isothermal amplification of substrates compete with each other. The higher binding affinity conferred by the canonical PAM allows cis-cleavage to dominate over amplification, leading to delayed or reduced amplicon generation. In contrast, adjusting the cleavage kinetics by employing a suboptimal PAM for Cas12a (Fig. 3b) can significantly enhance the assay’s detection efficiency due to reduced cis-cleavage during amplicon production. In this study, we hypothesized that applying the cis-cleavage-free SCas12a system (Fig. 3c) to DNA targets without PAMs would markedly improve the sensitivity, speed, and reproducibility of one-pot detection.

To validate the hypothesis, we first employed the one-pot RPA/SCas12a assay to detect the L1 gene of HPV16 (Fig. 3d) and the Orf1ab gene of SARS-CoV-2 (Fig. 3f), respectively. Meanwhile, WT-Cas12a targeting the identical DNAs but containing the canonical PAM (TTTA) was tested for comparison (Fig. 3e and g). The results showed that, within the first 10 min, the WT-Cas12a exhibited minimal fluorescence signal production. In contrast, the SCas12a exhibited an immediate increase in fluorescence, indicating no interference between RPA and Cas12a during the detection process. Consequently, the RPA/SCas12a assay demonstrated a faster reaction speed and generated significant fluorescence signals over a broader range of target concentrations. This conclusion is further supported by comparative analysis of amplicon profiles generated in the one-pot CRISPR assay using WT Cas12a and SCas12a. As shown in Supplementary Fig. S3, amplification products in the WT Cas12a system exhibit negligible accumulation within 15 min, which is consistent with efficient degradation mediated by its robust cis-cleavage activity. In contrast, the SCas12a system shows rapid and pronounced amplicon accumulation beginning at 5 min post-reaction initiation—confirming its lack of cis-cleavage activity.

The detection limits of the WT-Cas12a using the canonical PAM were 10 aM (∼6.0 copies/μl) for L1 and 100 fM (∼6.02 × 10⁴ copies/μl) for Orf1ab, respectively. In contrast, the detection limits achieved by the SCas12a were 0.1 aM (0.06 copies/μl) for L1 and 10 aM (∼6.0 copies/μl) for Orf1ab, representing a 100- to 1000-fold improvement in sensitivity, along with a significantly enhanced signal-to-noise ratio (Fig. 3d–g). Furthermore, the reliability of the SCas12a-based assay with PAM-free recognition and the WT-Cas12a-based assay with canonical PAM was evaluated by repeating the experiments under identical conditions ten times, with two replicates per run (Supplementary Fig. S4). The data demonstrate that the fluorescence signals from the SCas12a group were highly consistent across all replicates, whereas those from the WT-Cas12a group exhibited more than a 10-fold variation between replicates.

To broaden the application scope of the SCas12a assay, we further assessed its performance in a LAMP-mediated one-pot detection reaction using the thermostable CtCas12a at its optimal enzymatic temperature of 55°C. The E gene of SARS-CoV-2 was selected as the target, and the results demonstrated that the combination of LAMP with SCas12a exhibited significantly accelerated reaction kinetics compared to that with WT-Cas12a (Fig. 3h–i), achieving a 10-fold increase in sensitivity. These findings suggest that SCas12a markedly reduced interference with the LAMP process.

To validate the practical application of the one-pot testing method, we first employed the RPA/SCas12a assay to detect crude DNA extracted from 25 human vaginal secretion samples (Fig. 4a), which had previously been analyzed for HPV16 infection using qPCR (Supplementary Fig. S5). The method accurately identified all 12 HPV16-positive individuals among these clinical samples (Fig. 4b), demonstrating 100% sensitivity and 100% concordance with the qPCR results (Fig. 4c). In contrast, the one-pot CRISPR assay employing the classical DETECTR [10] method with WT Cas12a demonstrated a sensitivity of only 66.7%, detecting 8 out of the 12 samples. In conclusion, these findings demonstrate that SCas12a outperforms DETECTR in one-pot DNA detection assays.

To further verify the applicability of one-pot testing for pathogen-induced infections, we assessed the performance of RT–RPA/SCas12a using clinical specimens from individuals infected with SARS-CoV-2. The complete testing workflow is illustrated in Fig. 5a and requires ~40 min, comprising 5–10 min for viral inactivation and lysis by heating, followed by a 30-min RT–RPA/SCas12a reaction. The E and Orf1ab genes of SARS-CoV-2 were specifically detected (Fig. 5b) by SCas12a and WT Cas12a, with the former demonstrating at least a 10-fold higher detection efficiency. Subsequently, we applied the RT–RPA/SCas12a assay (Fig. 5c) to fifteen nasal swab specimens collected from individuals confirmed to have SARS-CoV-2 infection via clinical RT-qPCR analysis. These samples exhibited Ct values ranging from 11.59 to 27.23 (Supplementary Fig. S6). All fifteen samples were correctly identified as positive, showing a clear distinction in amplification slopes compared to three negative control swabs, resulting in a sensitivity of 100% (Fig. 5d). In contrast, the RT–RPA/WT-Cas12a achieved a sensitivity of 80%, detecting only 12 out of the 15 samples. Collectively, these findings demonstrate that SCas12a outperforms WT-Cas12a in one-pot diagnostics for RNA detection.

We and other groups have demonstrated that the SCas12a system exhibits significantly enhanced specificity for mutations located in the PAM-distal regions of DNA targets [36, 39, 40]. Compared to DNA with PAMs, the split crRNA is expected to show reduced affinity toward PAM-free DNA substrates, thereby amplifying the energetic penalty for single-nucleotide mismatches through nonequilibrium hybridization-driven regulation. Accordingly, we harnessed the SCas12a to test fully complementary dsDNA targets with single-nucleotide mismatches at positions 1–20 at a concentration of 1 nM (Fig. 6a). Compared to DNA subsrtate with a PAM, an obviously enhanced specificity was observed across positions 1 to 17, except for positions 10 and 19 (Fig. 6b). The highest DF for this assay was 91-fold, which indicates excellent single-nucleotide specificity in dsDNA detection.

The high single-nucleotide specificity makes SCas12a a powerful tool for rapid human genotyping. In this study, we validated the practical utility of the RPA/SCas12a assay in detecting the cancer-associated TP53 (rs215029805) SNP (A > T) [41] (Fig. 6c). We benchmarked the one-pot detection assay against genotyping results obtained from Sanger sequencing, which is the established gold standard for SNP indentification. Specifically, we obtained nine tissue samples from OSCC patients (seven WT and two A > T mutant samples) and applied the RPA/SCas12a system for SNP detection. The results showed a statistically significant difference in signal intensity between the WT and mutant samples (Fig. 6e). To further validate the reliability of the method, we introduced mutant-specific spacers for reverse confirmation of the mutant samples. These results demonstrated 100% concordance with Sanger sequencing data (Supplementary Fig. S7). The entire one-pot detection process could be completed within 30 min, which is substantially faster than SHERLOCK¹¹ (>2 h). Moreover, we established a standard for discriminating WT (DF > 4) and mutant type (DF < 0.25) using DF values.

The data underlying this article are available in the article and its online supplementary material.