CRISPR/Cas12a DTR system: a topology-guided Cas12a assay for specific dual detection of RNA and DNA targets

For this study, clinical human OSCC resection specimens for miRNA detection were collected by the Department of Oral and Maxillofacial Surgery at the First Affiliated Hospital of Zhengzhou University, with approval from the ethics committee.

6-Carboxyfluorescein (FAM)-labeled ssDNA and RNA probes, as well as all the DNA and RNA oligonucleotides (–), were synthesized by Sangon Biotech (Shanghai, China). Buffer preparation reagents were purchased from Sinopharm Chemical Reagent Co., Ltd (Beijing, China). NEB Buffers 1.1, 3.1, 4, and CutSmart were supplied by New England Biolabs (Beijing, China). The miRCURY LNA RT Kit was obtained from Qiagen, while SuperScript™ IV Reverse Transcriptase was acquired from Invitrogen™. RNase H, RNAPol buffer, ribonucleotide mix, murine RNase inhibitor, and T7 RNA polymerase were also sourced from NEB. DNase I-XT for RPA amplicon treatment was obtained as required. Proteinase K (Takara, Dalian, China) was used for protein degradation. Ultrapure water was utilized in all experiments. All DNA/RNA samples were dissolved in Diethylpyrocarbonate (DEPC)-treated water and stored at −20°C. Supplementary Tables S1 S2

Synthetic genes encoding AsCas12a and LbCas12a were cloned into the pET-28a(+) vector. Plasmid integrity was confirmed by restriction enzyme digestion and sequencing prior to transformation into Escherichia coli BL21 (DE3). A single colony was cultured in LB medium with 100 μg/ml kanamycin overnight at 37°C. The next day, the culture was diluted to OD₆₀₀ = 0.8 in 1 l of Terrific broth (TB), chilled on ice for 10 min, and induced with 0.5 mM Isopropyl-beta-D-thiogalactopyranoside (IPTG) at 18°C for 16 h. Cells were harvested by centrifugation, lysed in buffer A (20 mM Tris–HCl, pH 7.5, 300 mM NaCl, 1 mM PMSF, 5 mM β-mercaptoethanol) with protease inhibitors, and purified using Ni-NTA IMAC followed by size-exclusion chromatography on a HiLoad^® 16/600 Superdex^® column. Purified proteins were concentrated with a 100 kDa MWCO filter, quantified by Bradford assay, flash-frozen in liquid nitrogen, and stored at −80°C.

Initially, the Cas12a DTR systems were constructed in reaction buffer by mixing one of the Cas12a nucleases with its corresponding split crRNA at an equimolar ratio. This was followed by a 20-min incubation at room temperature. Subsequently, the trans-cleavage activity of the reconstituted Cas12a ribonucleoprotein (RNP) complex was assessed under the following conditions: a 10 μl reaction mixture comprising 1 μl of 10× NEB buffer 2.1, 40 nM Cas12a, 40 nM split RNA (20 nM miRNA-1 and 20 nM miRNA-2), and 20 nM of a 42-nucleotide ssDNA reporter in the presence of 40 nM ssDNA activators (20 nM ssDNA-1 and 20 nM ssDNA-2). The reactions were incubated at 37°C for 60 min and then terminated by heating at 90°C for 10 min. The resulting products were resolved on a 12% polyacrylamide gel (Fig. 2D), using 1× Tris-Borate-EDTA (TBE) as the electrophoresis buffer at a constant voltage of 100 V for 120 min. Finally, the gel was stained with GoldView and visualized under a UV transilluminator.

The DTR assay was carried out on the CFX96 Touch Real-Time PCR System (Bio-Rad, CA, USA) for fluorescence detection. Cas12a–crRNA complexes were prepared by mixing 500 nM Cas12a RNP with 1× NEB buffer 4, nuclease-free water, and 1 μM Polyvinyl pyrrolidone (PVP), followed by a 10-min incubation at room temperature. Auxiliary DNA activators (ssDNA, pseudo-hybrid DNA, and dsDNA) were subsequently added to the complexes. Each reaction consisted of 1 μM Fluorescence Quencher (FQ) reporter, target RNA at the appropriate concentration, and had a total volume of 20 μl. Reactions were incubated at 37°C for 20 min, and changes in fluorescence at 520 nm were monitored. Additionally, the same assay was performed using Lateral flow assay (LFA) detection (GenDx Cas12a detection kit, Suzhou, China).

Total RNA was extracted from tumor resection specimens of seven oral cancer patients and three healthy controls using the RNAprep Pure Micro Kit (TIANGEN, Beijing, China). The RNA was then used in a 20 μl DTR assay reaction to detect miR-155 and miR-let-7a. For Reverse transcription polymerase chain reaction (RT-qPCR) analysis, stem-loop primers and the Hairpin-it miRNA Detection Kit (GenePharma, Suzhou, China) were employed. RNA quantification followed the manufacturer's protocol, and a standard curve was generated to correlate miRNA concentrations with Cq values. Levels of miR-155 and miR-let-7a in clinical samples were determined using this curve. RT-qPCR was performed on a CFX96 Touch Real-Time PCR System (Bio-Rad, USA) with SYBR Green detection. Thermal cycling included initial denaturation at 94°C for 3 min, followed by 40 cycles of 94°C for 12 s and 62°C for 30 s.

The detection limit (LoD) was established by performing the trans-cleavage assay with different concentrations of RNA or DNA targets. The LoD value was computed using the formula 3σ/slope, in which σ represents the standard deviation of three blank measurements.

No statistical method was used to predetermine the sample size. All data were incorporated into the analyses without exclusion. The experiments were not randomized. Moreover, the investigators were informed of the allocation during the experiments and outcome assessment.

Recently, we have developed a split crRNA-based Cas12a detection method termed SCas12a for the direct detection of RNA targets [9]. However, this assay is not effective for simultaneous detection of two RNA targets in one pot (Fig. 2A), as either of them can initiate the trans-cleavage activity of Cas12a. To overcome this issue, we first designed a topology-guided Cas12a system, termed DTR, for noncontiguous target RNA activation (Fig. 2A). According to the design, the two target miRNAs are intended to function as spacer RNAs. Each of them comprises two ssDNA activator binding sites and a 2-nt bumper domain. This structural arrangement is expected to enable the simultaneous activation of both Cas12a nucleases.

For proof-of-concept study, we selected miR-let-7a and miR-155 as Spacer-1 and Spacer-2, which consists of 22 and 24 nucleotides, respectively. To investigate whether the break sites of the two spacers influence the catalytic activity of Cas12a, we designed nine combinations in which the break sites ranged from 6 to 14 (Fig. 2B). Additionally, the SCas12a detection system targeting these two miRNAs was employed as a control group (Fig. 2A). The fluorescence kinetic assays revealed that the combination (T8 + T12′) achieves the highest activation, exhibiting performance comparable to that of the SCas12a assay (Fig. 2C). The electrophoretic gel analysis further confirmed the effectiveness of the T8 + T12′ group (Fig. 2D). In addition, our results indicate that the presence of a single spacer (Fig. 2E, Control 1–1 and Control 1–2) is insufficient to activate the DTR system (Fig. 2F). Furthermore, the inclusion of a single activator (Fig. 2E, Control 2–1 and Control 2–2) results in only partial activation of the DTR system (Fig. 2F). These results provide evidence to support the notion that the Cas12a DTR system functions as a dimer, as proposed in Fig. 1A. Interestingly, the T12 + T8′ group, which exhibits the same break length but involves cleavage sites on distinct RNA substrates, demonstrates only marginal increases in fluorescence relative to the DTR system (Fig. 2C). This finding further suggests that the correct assembly of functional Cas12a dimer RNPs is critical for the DTR system.

To exclude potential sequence bias, we next examined various miRNA combinations (). Consistent results were obtained, indicating the universality of this assay for dual miRNA detection. Moreover, we dissolved the completely assembled Cas12a DTR system and checked it by 120 KeV Cryo-electron microscopy (cryo-EM). Many particles contain two Cas12a nucleases positioned next to each other (), with a first-to-first architecture, indicating the formation of complex as anticipated. In contrast, the cryo-EM images of Control 2–1 or Control 2–2, which contain one set of Cas12a RNPs, show much smaller sizes and exhibit single-protein morphology (). Finally, we conducted Dynamic Light Scattering (DLS) experiments, and the data () reveal that a significantly larger complex is formed in the DTR system compared to the Cas12a protein alone and the complex in Control 2–1. Supplementary Fig. S1 Supplementary Fig. S2A Supplementary Fig. S2B Supplementary Fig. S3

Collectively, these results demonstrate that the precise topological reconstitution of two active Cas12a complexes is essential for the functionality of the DTR detection system.

To determine whether Cas12a activity in the DTR system depends on the cooperative recognition of both targets, we established concentration gradients for miRNA substrates (T1 and T2) and evaluated their synergistic activation effects (Fig. 3A). Specifically, we anchored 10 nM of T1 and paired it with progressively decreasing concentrations of T2 (10 nM, 5 nM, 1 nM, 500 pM, and 100 pM). As the concentration of T2 decreased, the fluorescence rate markedly diminished, becoming negligible at T2 concentrations below 500 pM (Fig. 3B). A comparable trend was observed when the roles of T1 and T2 were interchanged, confirming that both components are indispensable for efficient activation (Fig. 3C). Furthermore, as a control, we tested equal concentrations of T1 and T2 ranging from 10 nM to 100 pM, which elicited a significantly stronger response within the same concentration groups (Fig. 3D). The results demonstrate that in the context of synergistic activation, the enzymatic activity of Cas12a is modulated by the combined concentrations of both activators rather than being strictly dictated by the lowest concentration of a single activator, indicating a concentration threshold and cooperative regulatory behavior (Fig. 3E). These findings emphasize the critical role of topological assembly and dual-input dependency in the DTR system, laying a foundation for downstream logic operations and experimental designs.

Boolean logic can process various types of information represented by 0 (NO) and 1 (YES), making it highly valuable for qualitative analysis [20, 21]. Here, we anticipated that the designed Cas12a DTR system could directly integrate input signal recognition, signal decoding, and effector output, enabling precise AND-logic detection of dual nucleic acid inputs (Fig. 4A). Firstly, we optimized logic gates specifically for miRNA and ssDNA detection. As shown in Fig. 4B, the two target miRNAs were designated as Input-1 and Input-2. The results demonstrated that Cas12a could only be effectively activated when all inputs precisely matched the supplied ssDNA activators, thereby forming a fully assembled closed-loop structure. In cases of absent or incomplete input signals, Cas12a activation was significantly suppressed due to incomplete structural assembly. Moreover, this strategy can be also applied for specific dual detection of ssDNA molecules when they serve as target nucleic acid inputs by providing the corresponding spacer RNAs. The DTR system is primarily designed for specific AND-logic detection; however, it can also be adapted for OR-logic detection of nucleic acids. For instance, in the detection of two miRNAs (Supplementary Fig. S4), two complementary activator ssDNAs were used instead of the spacer RNAs shown in Fig. 4B. Under these conditions, the presence of either Input-1 (miRNA-1) or Input-2 (miRNA-2) is sufficient to trigger Cas12a's trans-cleavage activity.

Recently, we developed the "Pseudo hybrid DNA-RNA" (PHD) assay by using a protospacer adjacent motif (PAM)-containing hybrid DNA activator for direct RNA sensing by Cas12a [22]. In this work, we utilized the same hybrid DNA activator (Fig. 4C) for the specific dual detection of two miRNA inputs, achieving a higher kinetic rate compared to the ssDNA activator. This design offers an additional efficient strategy for subsequent experimental planning. In the next step, we examined the applicability of the DTR system for dual dsDNA target detection. By using two independent dsDNA substrates as inputs (Fig. 4D), we evaluated their impact on Cas12a activation efficiency. The results demonstrated that, despite a slight reduction in kinetic rate caused by the inherent structural redundancy of dsDNA, the DTR system still reliably detected dual dsDNA targets. Finally, we successfully extended the DTR system to the AND-gate logic detection of RNA and DNA substrates (Fig. 4E). Through the rational design of activator and target nucleic acid sequences, we achieved, for the first time, the specific simultaneous detection of ssRNA and dsDNA.

Collectively, these findings highlight the advantages of the DTR system in integrating diverse nucleic acid signals and enabling topologically driven multitarget recognition and cooperative activation, thereby emphasizing its broad applicability and potential in nucleic acid multiplex detection.

We first evaluated the detection sensitivity of the DTR system for miRNA targets using the two strategies shown in Fig. 4B and C. By maintaining the ratio of the two miRNAs at 1:1 and varying their total concentration, the LoD of the DTR system was measured to be 78 fM (Fig. 5A) for the ssDNA activator and 767 fM for the hybrid DNA activators (Fig. 5B). The femtomolar-level sensitivity of the DTR system is comparable to that of previously reported Cas12a-based systems for miRNA detection [23–26]. Because ssDNA is simpler and more cost-effective, we adopted the first design strategy (Fig. 4B) for subsequent miRNA detection studies.

Next, we aimed to test more biologically relevant situations; i.e. the two target miRNAs usually exist in different ratios. We found that the DTR assay could significantly distinguish samples containing two miRNA targets with varied concentrations (1, 10, and 20 pM) and ratios (1:10 and 10:1) from negative controls and samples containing only one RNA target. However, no significant difference was observed when the concentrations of the two RNA targets were both ≥1 pM, regardless of their ratio (Fig. 5C). Additional experiments conducted at varying ratios further supported this conclusion (Supplementary Fig. S5), although the fluorescence intensity was lower at higher concentrations. These results confirmed the previously demonstrated synergistic activation of the DTR system and laid the foundation for clinical sample testing.

Beyond sensitivity, target recognition specificity is an even more critical criterion for evaluating nucleic acid detection methods. Given the stringent requirements for the assembly of components in the DTR system, we hypothesized that the DTR assay might exhibit superior specificity toward SNVs. To evaluate the specificity of the DTR assay for RNA substrates, we maintained miRNA-2 with constant sequences and designed a series of 20-nt target sequences in miRNA-1, each incorporating a single mismatch relative to the fully complementary sequence (Fig. 5D). The specificity of target recognition was evaluated by calculating the fluorescence ratio between mismatched and fully matched targets. A lower ratio indicates better discrimination of mismatched targets. Unexpectedly, the results demonstrated that the DTR system maintains high specificity for SNVs at all tested detection sites, with the exception of M1 (Fig. 5E). In contrast, the previously reported SCas12a assay [9] only demonstrated significant performance at M7, M12, M16, and M18. Additionally, we have conducted experiments to monitor the specificity of the DTR assay for dsDNA targets. The results show that the DTR system (Supplementary Fig. S6) exhibits a specificity similar to that of the conventional Cas12a system, with only a slightly higher specificity observed at the M10–M14 sites.

Together, the DTR method has been demonstrated to be a highly sensitive and specific nucleic acid detection approach, with broad prospects for clinical applications.

Cas12a DTR system provides the possibility to detect two RNAs simultaneously in the clinic to increase the accuracy of the single readout. OSCC is one of the most common head and neck malignancies, and the adequacy of surgical margins plays a critical role in determining the prognosis of the cancer [27–29]. However, currently, there remains an urgent need for effective intraoperative rapid detection methods for assessing resected tumor tissues. Previous studies have demonstrated the high-level expression of miR-let-7a and miR-155 in OSCC patient tumor tissues [28], ranging from picomolar to nanomolar concentrations, which falls within the dynamic range of the Cas12a DTR system. In our study, RNA was first extracted and purified from OSCC resection specimens and subsequently detected using the DTR fluorescence method within 20 min (Fig. 6A). The results (Fig. 6B and C) show that the DTR method effectively distinguishes patients with OSCC from healthy individuals. Moreover, by applying commercially available Cas12a LFA strips to clinical samples, we successfully identified positive cases among healthy controls (Fig. 6D). Overall, these findings highlight the excellent sensitivity and specificity of the Cas12a DTR method for disease diagnostics.