Cas-regulation-targeting chimera enables selective and tunable control of CRISPR/Cas12a

Abstract

Selective and tunable regulation of clustered regularly interspaced short palindromic repeats (CRISPR)/Cas12a activity enables on-demand control, yet current strategies remain hindered by nonspecific regulation and limited tunability. Inspired by proximity effect, we present a Cas-regulation-targeting chimera (CasTAC) strategy that employs CRISPR RNA (crRNA) as a proximity mediator to carry phosphorothioate regulators to interfere with catalytic or recognition domains of Cas12a and consequently suppress its activity. This crRNA-induced proximity approach can effectively eliminate nonspecific interaction between phosphorothioate regulators and proteins within complex multi-enzyme systems, thereby enabling selective control over CRISPR/Cas12a activity. Furthermore, CRISPR/Cas12a activity can be finely tuned to different inhibitory levels by varying the number of phosphorothioate regulators. The CasTAC strategy also improves nuclease resistance and single-nucleotide discrimination, offering potential advances in the sensitivity of molecular diagnostics and the accuracy of gene editing. Notably, the CasTAC balances the kinetics of nucleic acid amplification and CRISPR cleavage, facilitating efficient product accumulation and resolving compatibility issues in one-pot assays. As a proof of concept, we develop a one-pot, one-step recombinase polymerase amplification–CasTAC assay that achieves over 1000-fold higher detection sensitivity than the conventional one-pot recombinase polymerase amplification−CRISPR/Cas12a assay. The CasTAC strategy provides a versatile framework for fine–tuning Cas activity and advances CRISPR technology toward refined and context-adaptable functionality.

Graphical Abstract

Graphical Abstract.

Introduction

The clustered regularly interspaced short palindromic repeats (CRISPR)-associated (Cas) proteins, guided by CRISPR RNA (crRNA) or single-guide RNA, enable precise recognition and cleavage of nucleic acids, effectively integrating the catalytic amplification capacity of proteins with the sequence programmability of RNA [1–3]. The advent of CRISPR/Cas technology represents a transformative advance in biotechnology, underpinning diverse applications ranging from gene editing and molecular diagnostics to transcriptional regulation and live-cell imaging [4–8]. However, as its applications continue to expand, the intrinsic limitations of current CRISPR/Cas systems increasingly fall short of meeting emerging biological and clinical demands. A promising yet challenging direction lies in regulating Cas activity. Incorporating secondary structures into crRNA, such as G-quadruplexes or hairpins, allows conditional activation of Cas proteins and improves the specificity of CRISPR/Cas systems [9–12]. Photo-caging crRNAs with photosensitive groups enables spatiotemporal control of cleavage events, addressing compatibility issues between nucleic acid amplification and CRISPR cleavage [13–15]. Nevertheless, secondary-structure-based designs provide only limited stepwise regulation and cannot achieve tunable control of Cas activity. Likewise, most current photo-caging strategies depend on external light stimuli and offer only unidirectional, irreversible OFF/ON switching.

Against this backdrop, modulation of Cas protein activity using heparin sodium or phosphorothioate-modified DNA has emerged as a promising approach to advance CRISPR technology toward fine-tuned control in molecular biology applications [16, 17]. These regulators typically compete with crRNA or target DNA for Cas binding through electrostatic interactions, thereby occupying the catalytic or recognition domains and inhibiting enzymatic activity. However, such approaches still lack precision in controlling Cas functions. Regulation by heparin sodium is driven by nonspecific electrostatic interactions rather than molecular recognition, resulting in poor specificity. Similarly, phosphorothioate-modified DNA interacts with Cas proteins in a sequence-independent manner and may bind other nucleases within multi-enzyme systems, leading to undesired nonspecific interactions and reaction interference. Overall, tunable and selective control of CRISPR/Cas12a remains hindered by limited tunability and nonspecific regulation, compromising its accuracy and efficiency in molecular biology applications.

Proteolysis targeting chimera (PROTAC) technology exemplifies a chemically induced proximity strategy that recruits target proteins to E3 ligases through bifunctional molecules, thereby enabling selective degradation and minimizing nonspecific interactions [18–21]. Inspired by this concept, we hypothesized that a similar proximity-based approach could be adapted to carry phosphorothioate regulators to Cas12a for selective modulation of its activity. Given the sequence- and structure-dependent specificity of Cas12a binding to crRNA, crRNA is envisioned as an ideal proximity mediator to guide phosphorothioate regulators toward Cas12a [22].

Herein, we present a Cas-regulation-targeting chimera (CasTAC) strategy designed to eliminate nonspecific regulation and improve tunability, thereby enabling selective and tunable control of Cas12a activity. In this approach, crRNA functions as a proximity mediator that carries phosphorothioate regulators near Cas12a, enabling them to occupy the catalytic or recognition domains of the enzyme and consequently suppress its activity. In contrast to existing regulation driven by nonspecific interactions, the CasTAC strategy relies on sequence- and structure-dependent recognition, allowing selective modulation of Cas12a activity in multi-enzyme systems. Furthermore, CRISPR/Cas12a activity can be finely tuned to diverse inhibitory degrees by simply varying the number of phosphorothioate regulators. Such selective and tunable control improves nuclease resistance and single-nucleotide discrimination, offering potential advances in molecular diagnostic sensitivity and gene-editing accuracy. Moreover, CasTAC balances the kinetics of nucleic acid amplification and CRISPR cleavage, facilitating efficient product accumulation and resolving compatibility issues in one-pot assays. Building on this principle, a one-pot, one-step recombinase polymerase amplification−CasTAC (RPA−CasTAC) assay was developed, achieving over 1000-fold improvement in detection sensitivity compared with the conventional one-pot RPA−CRISPR/LbCas12a assay. Collectively, the CasTAC concept establishes a generalizable framework for the selective and tunable regulation of Cas activity and advances CRISPR technology toward fine-tuned control in molecular biology applications.

Materials and methods

Materials and reagents

All DNA and RNA oligonucleotides were synthesized and purified by Sangon Biotech. Co. Ltd. (Shanghai, China). Primers for RPA and polymerase chain reaction (PCR) were designed using NCBI online primer design toolkits (https://www.ncbi.nlm.nih.gov/). Detailed sequence information is provided in Supplementary Table S1. Plasmids containing sequences of Human Papillomavirus 16 (HPV-16), Influenza A virus (H1N1), and Mycobacterium tuberculosis (MTB) genes were cloned into the pUC57 vector and synthesized by Sangon Biotech Co., Ltd. (Shanghai, China). Sequence information is provided in Supplementary Table S2. The LbCas12a protein and NEBuffer™ r2.1 were purchased from New England Biolabs (USA). The AsCas12a and FnCas12a protein were obtained from Tolo Biotech. (Chuzhou, China). RNase R, ammonium persulphate (APS), N,N,N',N'-tetramethylethylenediamine, Gel red nucleic acid dye, Beyo Red DNA loading buffer (6×), and RNase-free water were purchased from Beyotime Biotech. Co., Ltd. (Shanghai, China). 40% acrylamide/bis-acrylamide (29:1), 5× Tris-borate-EDTA (TBE) Buffer, and 2× TaqMan Fast quantitative PCR (qPCR) Master Mix were obtained from Sangon Biotech. Co., Ltd. (Shanghai, China). TSINGKE TSE030 1.1× T3 Super PCR Mix was obtained from Tsingke Biotech. Co., Ltd. DNA Ladder was purchased from Takara Biomedical Technology Co., Ltd. (Beijing, China). DNA Isothermal Rapid Amplification Kit was obtained from Amp-Future Biotech. Co., Ltd. (Jiangsu, China).

Preparation of dsDNA and liner dsDNA substrates

The target double-stranded DNA (dsDNA) substrates containing the HPV16, H1N1, and MTB genes were prepared by mixing the target strand (TS) and non-TS at a 1:2 molar ratio in 1× NEBuffer™ r2.1. The strands were annealed by incubation at 95°C for 5 min, followed by a gradual cooling to 25°C at a rate of 0.1°C/s. The liner dsDNA was amplified from the plasmid through PCR. The PCR procedure (98°C for 15 s, 60°C for 30 s, 72°C for 15 s, 35 cycles) was performed on a thermocycler.

Trans-cleavage assay of CRISPR/Cas12a system

The trans-cleavage assay of CRISPR/Cas12a was conducted in a 20 μl reaction mixture containing 1× NEBuffer™ r2.1. The ternary complex was first assembled by incubating 10 nM LbCas12a, 10 nM crRNA, 50 nM or 20 nM phosphorothioate-anchored nucleic acids, and 1 nM target dsDNA at 37°C for 20 min. The cleavage reaction was then initiated by adding single-stranded DNA (ssDNA) FQ reporter to a final concentration of 500 or 100 nM. Fluorescence was monitored in real time for 1 h at 37°C using a CFX Connect Real-Time PCR Detection System (Bio-Rad, USA), with readings taken every 30 s. All experiments were performed in triplicate.

Cis-cleavage assay of CRISPR/Cas12a system

The cis-cleavage assay for polyacrylamide gel electrophoresis (PAGE) analysis was performed in a 10 μl system containing 100 nM LbCas12a, 100 nM crRNA, 20 ng linear dsDNA, and 1× NEBuffer™ r2.1. Prior to the addition of linear dsDNA, LbCas12a and crRNA were pre-incubated at room temperature for 30 min to form the ribonucleoprotein (RNP) complex. The reaction was then conducted at 37°C for 60 min and terminated by heating at 65°C for 10 min. The cleavage products were subsequently analyzed using PAGE (12% PAGE) in 1× TBE buffer at 110 V for 1 h at room temperature. Finally, the gel was stained with Gel red and visualized using a GenoSens 2150 imaging system (Clinx, China).

Cis-cleavage kinetics study

For kinetic analysis of CRISPR/Cas12a cis-cleavage kinetics of CRISPR/Cas12a, kinetic constants were quantified via tracking time-dependent changes in dsDNA substrate concentration. The reaction system consisted of 1× NEBuffer™ r2.1, 10 nM LbCas12a, 10 nM crRNA, and serially diluted target dsDNA spanning concentrations of 1, 2, 5, 10, and 15 nM. Fluorescence was monitored in real time for 1 h at 37°C using a CFX Connect Real-Time PCR Detection System (Bio-Rad, USA), with readings taken every 30 s. All experiments were performed in triplicate.

To quantify the cleavage activity, fluorescence curves were transformed into activity scores by fitting traces to effectively first-order reaction kinetics as previously described [23]. With a certain amount of active Cas12a, the concentration of uncleaved reporter is expected to decrease exponentially, and the cleavage process can be represented by the following reaction:

where E^★ represents the concentration of functionally active Cas12a, U is the concentration of uncleaved dsDNA reporter substrate, and P corresponds to the concentration of cleaved fluorescent reporter product. We defined r as the second-order intrinsic rate constant of this cleavage reaction. The time-dependent concentration of cleaved fluorescent product P(t) follows the single-exponential saturation function:

where P_tot represents the fitted plateau fluorescence, P₀ is the initial fluorescence, and k_obs refers to the apparent activity rate constant. Under the assumption that E^★ is constant over time, k_obs = rE^★ is defined as effective first-order rate constant (units of min⁻¹). Assuming that r is constant across our assays, k_obs therefore serves as a relative measure of active Cas12a levels. Given measured P₀, nonlinear least-squares fitting was performed in MATLAB via a trust-region algorithm to determine the best-fit values of P_tot and k_obs.

PAGE analysis of pre-crRNA processing

To investigate whether psDNA affects the RNase activity of LbCas12a, 12% PAGE was employed to monitor the maturation of pre-crRNA. The LbCas12a-mediated pre-crRNA processing was performed in a 10 μl reaction mixture containing 1× NEBuffer™ r2.1, 200 nM LbCas12a, 200 nM crRNA, and 200 nM psDNA. A control reaction without psDNA (Ctrl.) was performed in parallel. Reactions were carried out at 37°C for varying time points (ranging from 0 to 60 min) and terminated by heating at 65°C for 10 min. The resulting mixtures were analyzed using 12% PAGE in 1× TBE buffer at 110 V for 1 h at room temperature. Finally, the gel was stained with Gel red and visualized using a GenoSens 2150 imaging system (Clinx, China).

Electrophoretic mobility shift assay

To assess the binding affinity crRNA, psDNA, and pscrRNA to LbCas12a, electrophoretic mobility shift assay (EMSA) was performed using LbCas12a protein at serially diluted concentrations ranging from 10 nM to 1 µM. Binding assays were performed in 1× TBE buffer with a constant nucleic acid concentration of 200 nM. After incubation at 37°C for 30 min, the samples were analyzed by 12% PAGE in 1× TBE buffer at 110 V for 1 h at room temperature. The gel was subsequently stained with Gel red and visualized using a GenoSens 2150 imaging system.

Molecular docking simulations

For complex model construction, both the protein structure and nucleic acid sequence were obtained from the Protein Data Bank (PDB), prioritizing entries with a resolution >2.5 Å and complete structural integrity. The Chai–1 multimodal foundation model, which supports the prediction of protein, small molecule, DNA, RNA, and glycosylation structures (see https://github.com/chaidiscovery/chai-lab), was then used to perform unified structure prediction and complex assembly for the protein and target nucleic acid fragment. Specifically, the protein backbone and nucleic acid base sequence were input into Chai–1, which generated initial complex coordinates and inferred interaction patterns between interfacial side chains and nucleotide bases. The output structure was locally refined using the OLPS4 force field in Maestro 13.0 to maintain plausible hydrogen bonding networks and hydrophobic contacts. Finally, energy minimization was applied to the resulting complex (convergence threshold 0.02 Å) to yield a stable conformation for subsequent binding energy calculations.

For MM-GBSA binding energy calculation, based on the optimized complex structure, MM-GBSA calculations were performed using the Prime module of the Schrödinger 2021–4 suite. The minimized structures of the complex, receptor, and ligand were extracted separately. Each of the three systems underwent local re-minimization under the OPLS3e force field for electrostatics and van der Waals interactions, along with the VSGB solvation model, using an energy convergence threshold of 0.05 kcal·mol⁻¹·Å⁻¹. The total binding free energy was then calculated as:

Trans-cleavage assays of CRISPR-nonspecific RPA system

The trans-cleavage assay of CRISPR-nonspecific RPA system was conducted in a 15 μl reaction mixture containing 1× NEBuffer™ r2.1. The ternary complex was first assembled by incubating 10 nM LbCas12a, 10 nM crRNA, and 1 nM target DNA (HPV-16) at 37°C for 20 min. Subsequently, a single RPA pellet was dispersed into 29.4 μl of rehydration buffer to form a suspension. Then, 5.9 μl of the resuspended RPA mixture, 0.3 μl of forward primer and 0.3 μl of reverse primer (10 μM) targeting the H1N1 gene, 1.5 μl of H1N1 plasmid, 0.54 μl of MgOAc (280 mM), and 100 nM FQ reporter were added sequentially. Fluorescence was monitored in real time for 1 h at 37°C using a CFX Connect Real-Time PCR Detection System (Bio-Rad, USA), with readings taken every 30 s.

Trans-cleavage assays of CasTAC system

The trans-cleavage assay of CasTAC system was conducted in a 20 μl reaction mixture containing 1× NEBuffer™ r2.1. The ternary complex was first assembled by incubating 10 nM LbCas12a, 10 nM pscrRNA, and the target DNA at 37°C for 20 min. The cleavage reaction was then initiated by adding ssDNA FQ reporter to a final concentration of 500 nM. Fluorescence was monitored in real time for 1 h at 37°C using a CFX Connect Real-Time PCR Detection System (Bio-Rad, USA), with readings taken every 30 s.

RNase R digestion assay

The RNase R digestion assay was performed in a 20 μl reaction comprising 2 μl of 10× RNase R reaction buffer, 1 μl of RNase R (2 U/μl), 10 μl of crRNA or pscrRNA (1 μM), and 7 μl of RNase-free water. The mixture was incubated at 37°C for various durations (0, 0.5, 1, 2, 4, 8, 12, and 24 h) and terminated by heating at 80°C for 10 min. The products were then analyzed by 12% PAGE assay or trans-cleavage assay.

Stability assay of CRISPR/Cas12a and CasTAC system

To compare the stability of the two systems, varying amounts of RNase R was introduced to mimic different levels of nuclease contamination. Apart from the addition of RNase R, all other reaction components and conditions were consistent with the trans-cleavage assay of CasTAC system described earlier.

Beacon assay

The Beacon assay was conducted in a 20 μl reaction mixture containing 1× NEBuffer™ r2.1. First, Beacon dsDNA was obtained through an annealing reaction. The 100 μl annealing reaction system comprised 1× NEBuffer™ r2.1, 10 µM each of Beacon-17nt, Beacon-FAM, and Beacon-BHQ ssDNA sequences. The strands were annealed by incubation at 95°C for 5 min, followed by a gradual cooling to 25°C at a rate of 0.1°C/s. The Beacon assay was conducted in a 20 μl reaction mixture containing 1× NEBuffer™ r2.1, 5 nM LbCas12a, 5 nM crRNA, and 50 nM Beacon dsDNA. Fluorescence was monitored in real time for 1 h at 37°C using a CFX Connect Real-Time PCR Detection System (Bio-Rad, USA), with readings taken every 30 s.

Conventional one-pot RPA−CRISPR/Cas12a assay

First, a single RPA pellet was resuspended in 29.4 μl of primer-free rehydration buffer to form a suspension. The reaction mixture was prepared with two separately components, A and B. Component A was prepared by combining 1.5 μl 10× NEBuffer™ r2.1, 0.3 μl LbCas12a (1 μM), 0.3 μl of crRNA (1 μM), 1.5 μl FQ probe (5 μM), 0.54 μl MgOAc (280 mM), and 0.86 μl RNase-free water. Component B consists of 0.3 μl each of forward and reverse primers (10 μM), 5.9 μl suspended RPA mixture, 1.5 μl plasmid, and 2 μl RNase-free water. Component B was first added to the bottom of the eppendorf (EP) tube. Components A was then added on the lid of the EP tube. After mixing components A and B, the mixture was incubated at 37°C for 90 min, and the fluorescence signal of the mixture was recorded with a CFX Connect Real-Time PCR Detection System.

One-pot, one-step RPA−CasTAC assay

The experiment was conducted under conditions identical to those described earlier, except for the replacement of crRNA with pscrRNA.

Real-time fluorescence monitoring of the RPA amplification

The DNA Isothermal Rapid Amplification Kit was used for RPA amplification. First, a single RPA pellet was resuspended in 29.4 μl of primer-free rehydration buffer to form a suspension. The RPA reaction was then performed in 15 μl mixture containing 5.9 μl suspended RPA mixture, 0.3 μl each of forward and reverse primers (10 μM), 1.5 μl plasmid, 0.54 μl MgOAc (280 mM), 1.5 μl SYBR Green (10×), and RNase-free water. To investigate the effect of pscrRNA on RPA amplification, the RPA reaction was carried out in a 15 μl mixture, with an additional 0.3 μl of pscrRNA (1 μM) being added. To investigate the effect of CasTAC system on RPA product accumulation, the RPA assay was conducted under the same conditions as the one-pot, one-step RPA−CasTAC experiment, except that the 1.5 μl FQ probe (5 μM) was changed to 1× SYBR Green. To investigate the effect of RPA−CRISPR/Cas12a system on RPA product accumulation, the RPA assay was conducted under the same conditions as the conventional one-pot RPA−CRISPR/Cas12a experiment, except that the 1.5 μl FQ probe (5 μM) was changed to 1× SYBR Green. The reactions were incubated at 39°C for 20 min, and the fluorescence signal of the mixture was recorded with a CFX Connect Real-Time PCR Detection System.

Agarose gel analysis of RPA product accumulation in the one-step assay

For the agarose gel analysis, the RPA reaction was then performed in a 15 μl mixture containing 5.9 μl suspended RPA mixture, 0.3 μl each of forward and reverse primers (10 μM), 1.5 μl plasmid (10 fg/μl), 0.54 μl MgOAc (280 mM), and RNase-free water. To assess the effect of RPA−CRISPR/Cas12a or RPA−CasTAC system on RPA product accumulation, the RPA experiment was conducted under the same conditions as the one-pot experiment, except that the plasmid was 10 fg/μl and did not contain the FQ probe. The reactions were incubated at 39°C, and samples were collected at five time points (0, 5, 10, 15, and 20 min). The amplification products were purified by phenol-chloroform extraction (25:24:1), followed by centrifugation at 12 000 rpm for 5 min. 5 μl of purified products were visualized on 1.5% agarose gel.

Detection of clinical DNA samples by qPCR

The clinical samples tested in this study were approved by the Ethics Committee of Chongqing Medical University (reference number: 2024128). A total of 105 MTB-suspected samples for this research were collected from patients who visited the First Affiliated Hospital of Chongqing Medical University for routine diagnosis and treatment. All participants provided informed consent prior to sample collection. The clinical samples were fully de-identified prior to analysis. The MTB nucleic acid detection Kit (ZJ Bio-Tech, No. Z-ME-0072) was used to extract MTB DNA from clinical samples in a biosafety level II laboratory, and the extracted MTB DNA samples were tested using the standard qPCR.

A 20 μl qPCR reaction mixture contained 2× TaqMan Fast qPCR Master Mix, primers (200 nM each), and a TaqMan probe (200 nM). To each reaction, 3 μl of DNA extracted from clinical samples was added. Fluorescence signals were recorded using an ABI Q3 instrument (Thermo Fisher Scientific). For MTB-suspected samples, the qPCR thermal cycling program was set as follows: pre-denaturation at 94°C for 3 min, followed by 40 cycles of amplification (94°C for 5 s and 60°C for 30 s). According to the manufacturer’s protocol of the kit, the criteria were claimed as: C_t value ≤ 40 (positive), C_t value > 40 (negative).

For the one-pot, one-step RPA−CasTAC assay of clinical samples, the testing procedure was performed as described previously, with the exception that the plasmids were replaced by the DNA extracts from clinical samples.

Statistics and reproducibility

The EMSA gels band intensities were quantified using ImageJ software (National Institutes of Health). The resulting data were then plotted and fitted with a nonlinear regression model in GraphPad Prism software (Version 10.1.2) to determine the dissociation constant (K_d). Statistical tests were performed using GraphPad Prism software for unpaired two-tailed Student’s t-test to evaluate statistical differences between two groups, where ns denotes non-significant with P > .05, and asterisks refer to significant differences with *P < .05, **P < .01, ***P < .001, and ^****P < .0001. All experimental results were derived from at least three biologically independent experiments, and data are presented as the mean ± standard deviation (mean ± s.d.).

Results

Regulation of Cas12a activity with phosphorothioate-anchored nucleic acids

The phosphorothioate modification of nucleic acids involves introducing sulfur atoms into phosphodiester bonds, which can significantly enhance the stability of nucleic acids and improve their protein binding ability [24–26]. Previous studies have reported that phosphorothioate-anchored nucleic acids can serve as potent inhibitors of the cis-cleavage activity of Cas12a [27]. However, the effect of phosphorothioate-anchored nucleic acids on the trans-cleavage activity, a distinctive feature of Cas12a that differentiates it from Cas9, warrants more comprehensive investigation. Elucidating the interaction between phosphorothioate-anchored nucleic acids and the CRISPR/Cas system may provide deeper insights into nucleic acid–protein interactions and facilitate the development of novel regulatory tools.

To systematically evaluate the regulatory effect of phosphorothioate-anchored nucleic acids on LbCas12a activity, a 41-nucleotide phosphorothioate-anchored DNA with a random sequence (psDNA41) was designed and synthesized [27]. Two crRNAs were designed to target the HPV-16 gene and H1N1 gene, respectively. To assess the trans-cleavage activity of LbCas12a, a FRET-based fluorescent reporter (FQ reporter) was employed, which generates enhanced fluorescence upon cleavage (Fig. 1A). Incubation of the LbCas12a/crRNA complex with the FQ reporter and increasing concentrations of target DNA generated progressively increased fluorescence signals (Fig. 1B and C; Supplementary Fig. S1), confirming LbCas12a activation and subsequent trans-cleavage of the reporter. However, the introduction of psDNA41 into the reaction markedly suppressed fluorescence to a level comparable to the blank control, indicating that psDNA41 almost completely inhibited the trans-cleavage activity of LbCas12a. When PCR-amplified HPV-16 and H1N1 gene fragments were used as substrates, psDNA41 similarly suppressed the cis-cleavage activity of LbCas12a (Fig. 1D–F and Supplementary Fig. S2). These findings indicate that psDNA exerts a strong inhibitory effect on both cis- and trans-cleavage activities of LbCas12a. Moreover, both cleavage activities were progressively reduced as the concentration or length of psDNA increased (Fig. 1G and H; Supplementary Figs S3–S5), demonstrating that psDNA enables tunable modulation of LbCas12a activity in a concentration- and length-dependent manner. To exclude the possibility that inhibition resulted from nonspecific competition with the ssDNA reporter, ssDNA without phosphorothioate regulators (DNA41) was tested under identical conditions. In contrast to psDNA41, unmodified DNA41 exhibited no detectable inhibitory effect (Supplementary Fig. S6), demonstrating that the inhibitory activity of psDNA41 originates from the phosphorothioate regulators. Furthermore, similar inhibitory patterns were observed across multiple LbCas12a orthologs, suggesting that this regulatory effect is broadly applicable within the Cas12a family (Supplementary Figs S7 and S8).

Figure 1.

Regulation of LbCas12a activity by phosphorothioate-anchored nucleic acids. (A) Schematic illustration of psDNA-mediated regulation of LbCas12a trans-cleavage activity. Evaluation of the effects of psDNA (50 nM) on the LbCas12a trans-cleavage activity using target dsDNA containing HPV-16 gene (B) and H1N1 gene (C) at three different concentrations. (D) Schematic illustration of psDNA-mediated regulation of LbCas12a cis-cleavage. Time-dependent accumulation of cis-cleavage products analyzed by 12% PAGE in the absence (E) and presence (F) of psDNA. The uncut control (UC) represents linear dsDNA amplified from the HPV-16 plasmid. Effects of psDNA concentration (G), length (H), sequence composition (I), types of phosphorothioate-anchored nucleic acids (J), and phosphorothioate modification distribution (K) on the LbCas12a trans-cleavage activity. The sequence of RAM-psDNA21 in Fig. 1I is consistent with that of psDNA21. All phosphorothioate-anchored nucleic acids used in Fig. 1H–K are at a concentration of 20 nM. The histograms show relative fluorescence units (RFU) measured at 60 min. Ctrl. represents CRISPR/Cas12a assay without phosphorothioate-anchored nucleic acids. PC represents positive control with target. NTC represents negative control without target. Two-tailed Student’s t-test; ^****P < .0001; ns, not significant. Data represent mean ± SD (n = 3 replicates).

These findings prompted us to explore additional parameters influencing phosphorothioate-mediated inhibition. To examine sequence preference, we compared the effects of random, AT-rich, and GC-rich psDNA on the trans-cleavage activity of LbCas12a. Across tests with two different targets, GC-rich psDNA21 exhibited the strongest inhibitory effect on the trans-cleavage of LbCas12a (Fig. 1I and Supplementary Fig. S9). Given the inherent structural, stability, and chemical differences between DNA and RNA, we further compared different types of nucleotides with phosphorothioate modifications. Compared with the corresponding psRNA21, random-sequence psDNA21 exhibited substantially stronger inhibitory activity against LbCas12a (Fig. 1J and Supplementary Fig. S10). In addition, we found that the inhibition efficiency was also influenced by the distribution of phosphorothioate on the DNA backbone. The psDNA containing closely spaced phosphorothioate modifications displayed higher inhibitory efficiency than sparsely modified counterparts (Fig. 1K and Supplementary Fig. S11). Overall, these results indicate that phosphorothioate-anchored nucleic acids enable tunable regulation of CRISPR/LbCas12a activity in a sequence-independent manner.

Mechanistic insights of the LbCas12a activity inhibition by phosphorothioate-anchored DNA

As shown in Fig. 2A, LbCas12a first binds to the repeat region of crRNA to form an RNP complex. This complex subsequently recognizes the target DNA to assemble into the LbCas12a/crRNA/DNA ternary complex. Upon complete hybridization of the crRNA to the TS, cis-cleavage activity and trans-cleavage activity are activated [28–30]. To clarify the mechanism by which psDNA inhibits LbCas12a activity, we designed a staged assembly strategy combined with trans-cleavage assays (Fig. 2B). We first investigated whether psDNA interfered with RNP formation. To this end, psDNA, crRNA, target DNA, and reporter were simultaneously added into a reaction system containing pre-incubated LbCas12a alone. Under this condition, negligible fluorescence signal was observed (Fig. 2C-i). PAGE analysis further indicated that psDNA impeded RNP formation (Supplementary Fig. S12). To exclude the possibility that inhibition solely originated from impaired RNP assembly, we next pre-incubated LbCas12a with crRNA to allow RNP formation prior to psDNA addition. Despite successful pre-assembly of the RNP complex, trans-cleavage activity remained strongly inhibited (Fig. 2C-ii). PAGE analysis further revealed a substantial reduction in ternary complex formation after psDNA introduction, suggesting that psDNA additionally interferes with target DNA binding and ternary complex assembly (Supplementary Fig. S13). We further asked whether psDNA could still inhibit LbCas12a after formation of the activated ternary complex. Therefore, the ternary complex was pre-assembled prior to psDNA addition to ensure that LbCas12a had already entered the activated cleavage state. Even under this condition, trans-cleavage activity was rapidly suppressed following psDNA introduction (Fig. 2C-iii). PAGE analysis showed that substantial cis-cleavage products had already accumulated within 30 min, indicating successful activation of the cleavage process before psDNA addition (Fig. 2D). Moreover, we extended the pre-incubation to 120 min to ensure the formation of more ternary complexes with trans-cleavage activity, but the fluorescence signal remained negligible (Fig. 2E). Moreover, introducing psDNA at any time point during cis-cleavage immediately prevented further accumulation of cis-cleavage products (Supplementary Fig. S14). This result indicates that even when ternary complexes with trans-cleavage activity have already formed, the introduction of psDNA also can rapidly inhibit trans-cleavage, suggesting that psDNA may directly act on the catalytic center under this condition.

Figure 2.

Mechanistic insights of psDNA-mediated regulation of LbCas12a activity. (A) Workflow of Cas12a-mediated DNA cleavage. (B) Schematic diagram of the reaction between psDNA and CRISPR/LbCas12a system under three different reaction conditions. i: pre-incubate LbCas12a individually, followed by simultaneous addition of psDNA, crRNA, target DNA and reporter. ii: LbCas12a was preincubated with crRNA for RNP assembly before psDNA addition. iii: the ternary complex was pre-assembled prior to psDNA addition. (C) Evaluation of the trans-cleavage activity of LbCas12a under three different reaction conditions. (D) 12% PAGE analysis of cis-cleavage yields as a function of incubation time for LbCas12a/crRNA/DNA ternary complex. (E) The influence of psDNA on the trans-cleavage activity of LbCas12a during different pre-incubation processes of LbCas12a/crRNA/DNA ternary complex. (F) Schematic diagram of LbCas12a-mediated pre-crRNA processing. (G) Time-dependent of mature-crRNA generation in the absence (top) and presence (bottom) of psDNA. Pre represents pre-crRNA alone. The binding affinity of LbCas12a/crRNA complex (H) and LbCas12a/psDNA complex (I). The left panel is 12% PAGE results, and the right panel shows the fitted curve. The concentrations of LbCas12a are 0, 10, 20, 50, 100, 200, 400, 600, 800, and 1000 nM respectively. (J) The structure of the LbCas12a protein. (K) The docking result of LbCas12a with crRNA, psDNA, or both crRNA and psDNA. (L) Comparison of binding energies for LbCas12a/crRNA and LbCas12a/psDNA complexes. psDNA41 was used for the experiments in this figure. Ctrl. represents CRISPR/Cas12a assay without phosphorothioate-anchored nucleic acids. NTC represents negative control without target. Data represent mean ± SD (n = 3 replicates).

Cas12a harbors independent DNase and RNase activities, enabling cleavage of both DNA and RNA substrates [22, 31]. Beyond canonical substrate cleavage, Cas12a is capable of processing precursor crRNA (pre-crRNA) upstream of stem-loop structure (Fig. 2F). Unlike the RuvC domain responsible for DNA cleavage, pre-crRNA processing occurs within the WED domain [31, 32]. We therefore hypothesized that psDNA may also interact with the WED domain, thereby impeding the processing of pre-crRNA. The mature crRNA migrates faster than pre-crRNA with high-molecular weight in gels (Fig. 2G). After the addition of psDNA, we observed a significant reduction in the bands corresponding to mature crRNA. These findings further support our hypothesis that psDNA not only inhibits the DNase activity of LbCas12a but also markedly suppresses its RNase activity.

We found that psDNA can rapidly inhibit LbCas12a activity even after LbCas12a activation, suggesting a stronger binding affinity between psDNA and LbCas12a. To evaluate this interaction, we performed EMSA using crRNA and a psDNA of equivalent length (psDNA41). The upper gel band corresponds to the LbCas12a/nucleic acid complex, while the unbound crRNA and psDNA migrate as the lower gel band (Fig. 2H and I, right panel). With increasing LbCas12a concentration, the binding of both nucleic acids to LbCas12a was enhanced. Fitting of the binding curves revealed that psDNA binds to LbCas12a with ~3.8-fold higher affinity than crRNA (Fig. 2H and I, left panel).

To further investigate the mechanism underlying psDNA-mediated regulation of LbCas12a, we performed molecular docking simulations to compare the binding conformations and interactions of LbCas12a with psDNA versus crRNA. The structure of LbCas12a (PDB: 5ID6) used for docking is shown in Fig. 2J. Previous studies have indicated that the broad distribution of crRNA across the Cas12a surface is essential for subsequent catalytic processes [31, 33]. In the LbCas12a/crRNA binary system, crRNA engaged extensively with multiple domains, including the recognition lobes (WED-I, REC1, and REC2) and catalytic regions (RuvC-II, Nuc, and RuvC-III) (Fig. 2K, left panel and Supplementary Fig. S15A). These interactions help stabilize crRNA in the catalytic cavity and may facilitate DNA targeting and cleavage. Similarly, psDNA was found to bind key regions spanning from the nucleic acid recognition region to the catalytic core (Fig. 2K, middle panel). Specifically, psDNA occupies sites near the crRNA channel entrance and initial recognition surface in WED-I, REC1, and REC2, suggesting deep penetration into the nucleic acid binding channel. Notably, psDNA also closely contacts with catalytically critical regions including WED-III, RuvC-I/II/III, and Nuc. Binding within RuvC-I and RuvC-II implies potential structural perturbation of the catalytic pocket, thereby inhibiting enzymatic activity (Supplementary Fig. S15B). To further distinguish the binding modes of crRNA and psDNA, we performed competitive docking. Structural analysis revealed that psDNA displaced crRNA from its canonical binding sites (Fig. 2K, right panel), indicating competitive inhibition of LbCas12a/crRNA complex formation. The theoretical binding energies were –150.40 kJ·mol⁻¹ for LbCas12a/crRNA and –408.39 kJ·mol⁻¹ for LbCas12a/psDNA, implying that psDNA exhibited markedly higher affinity, in agreement with our experimental results (Fig. 2L). Collectively, the staged assembly experiments and molecular docking analyses support a multi-stage inhibitory mechanism of psDNA toward LbCas12a. These results suggest that psDNA can competitively interfere with crRNA loading, disrupt ternary complex formation, or directly perturb the catalytic regions of activated LbCas12a, depending on the stage at which it is introduced.

Selective and tunable regulation of LbCas12a activity in multi-enzymes system

Phosphorothioate-anchored DNA exhibits broad protein affinity, which may result in unintended binding to other nucleases within multi-enzyme systems. To evaluate whether psDNA can specifically regulate LbCas12a activity under such conditions, we established a one-pot RPA-CRISPR/LbCas12a system. This system integrates multiple enzymes, including recombinase, DNA polymerase, recombinase-loading factors, ssDNA-binding proteins, and creatine kinase, thereby providing a relevant environment to assess psDNA-mediated regulation of Cas12a. The fluorescence signal remained comparable to that of the control group (without psDNA41) even with increasing psDNA41 concentrations, indicating that psDNA41 failed to regulate LbCas12a activity in the one-pot system (Supplementary Fig. S16). Given that psDNA exhibited strong binding affinity toward LbCas12a, we further investigated whether pre-incubation of psDNA with LbCas12a could selectively modulate LbCas12a activity and thereby improve the sensitivity of the one-pot assay. As shown in Supplementary Fig. S17, pre-incubation of 50 nM psDNA41 with LbCas12a for 5, 10, 20, and 30 min showed no obvious enhancement in fluorescence intensity compared with the conventional one-pot assay. Moreover, the fluorescence signal gradually decreased with the increasing psDNA41 concentration, possibly because higher concentrations of psDNA41 nonspecifically interfered with proteins involved in the RPA amplification process. These findings indicate that pre-assembly of free psDNA41 with LbCas12a cannot significantly achieve effective regulation of Cas12a activity in the complex multi-enzyme RPA system. To further clarify the influence of RPA components, we designed a nonspecific RPA system whose amplification product was non-homologous to the crRNA, and integrated it with CRISPR/LbCas12a into a one-pot system. While psDNA41 markedly suppressed LbCas12a activity in the CRISPR-only reaction, this inhibition was completely abolished in the presence of RPA components (Fig. 3A and Supplementary Fig. S18). Given that both recombinase and ssDNA-binding proteins exhibit strong affinity for ssDNA, they likely sequester psDNA, rendering it inaccessible to LbCas12a. To verify this, psDNA was annealed to form double-stranded psDNA (ds-psDNA). We observed similar results to those of psDNA. (Fig. 3B and Supplementary Fig. S19). These findings suggest that psDNA fails to exert specific inhibition on LbCas12a activity in multi-enzyme systems. Such limitations may lead to undesired nonspecific binding and reaction interference, thereby restricting its applicability in complex biological environments.

Figure 3.

Selective regulation of LbCas12a activity in multi-enzyme system. Real-time fluorescence curves showing the regulation of LbCas12a trans-cleavage activity by psDNA (A), ds-psDNA (B), crRNA-psDNA (C), or pscrRNA (D) in the CRISPR-only (top) and CRISPR-nonspecific RPA (bottom) systems. The FQ-reporter concentration is 100 nM. (E) Schematic diagram of crRNA spacer modifications with different numbers of phosphorothioate linkage. (F) CRISPR/LbCas12a trans-cleavage assay using crRNAs modified with different numbers of phosphorothioate linkage. The histograms show RFU measured at 60 min. (G) One-pot RPA-CRISPR/LbCas12a assay using crRNAs modified with different number of phosphorothioate. The histograms show RFU measured at 90 min. (H) Docking calculations of LbCas12a bound to crRNA (right) and pscrRNA (left). The pscrRNA20 was used in this experiment. (I) Comparison of binding energies between the LbCas12a/crRNA and LbCas12a/pscrRNA complexes. (J) The binding affinity of LbCas12a/pscrRNA20 complex. The left panel is PAGE results, and the right panel shows the fitted curve. The concentrations of LbCas12a are 0, 10, 20, 50, 100, 200, 400, 600, 800, and 1000 nM respectively. Ctrl. represents CRISPR/Cas12a assay without phosphorothioate-anchored nucleic acids. PC represents positive control with target. NTC represents negative control without target. Data represent mean ± SD (n = 3 replicates).

The nonspecific regulation observed in multi-enzyme environments likely originates from nonspecific interactions between psDNA and other proteins. Therefore, minimizing such nonspecific binding to achieve selective regulation of Cas12a represents a key challenge. PROTAC technology minimizes nonspecific interactions by forcibly recruiting the target protein to the E3 ligase via bifunctional molecules [19]. Considering that Cas12a specifically recognizes the repeat region of crRNA in a structure- and sequence-dependent manner [22], we envisioned that crRNA could serve as a molecular carrier to localize phosphorothioate regulators near LbCas12a, enabling selective and tunable control of its activity. We term this mechanism “crRNA-induced proximity,” which inspired the development of the CasTAC strategy. As a proof of concept, a 16-nt psDNA was conjugated to the 3′-end of the crRNA (crRNA-psDNA16). A crRNA-DNA conjugate (crRNA-DNA16) without phosphorothioate served as the control. As shown in Fig. 3C and Supplementary Fig. S20, extension of the crRNA with unmodified DNA at the 3′ end enhanced the trans-cleavage activity of LbCas12a in both CRISPR-only and CRISPR-nonspecific RPA systems. In contrast, conjugation with psDNA at the 3′ end markedly suppressed LbCas12a activity. Given the observed inhibitory effects of psRNA, we subsequently introduced phosphorothioate groups directly into the spacer region of crRNA (pscrRNA) and combined it with LbCas12a to establish the CasTAC system. Compared with the conventional CRISPR/LbCas12a system employing unmodified crRNA, the CasTAC system exhibited a significantly reduced fluorescence signal (Fig. 3D and Supplementary Fig. S21), indicating that phosphorothioate regulators positioned in proximity to LbCas12a can selectively modulate its catalytic activity. Moreover, selective regulation was observed for the two homologous proteins, highlighting the versatility of the design (Supplementary Fig. S22).

Next, we systematically explored the regulatory effects of phosphorothioate regulators introduced at different sites of crRNA on enzymatic activity. As shown in Supplementary Fig. S23A, the crRNA was divided into four regions (A, B, C, and D) for phosphorothioate modification. The results demonstrated that phosphorothioate regulators in each of the above four regions exhibited no significant differences in the regulatory effects on the trans-cleavage activity of LbCas12a (Supplementary Fig. S23B and C). Given that the repeat region of crRNA is mainly responsible for its specific binding to LbCas12a, the spacer region lies closer to the catalytic center and better suited for selective modulation of Cas12a activity. Therefore, we focused on investigating the regulatory effects of phosphorothioate modifications within the spacer region. We systematically evaluated the effects of phosphorothioate regulators on both the cis- and trans-cleavage activities of the CRISPR/LbCas12a system, and observed that both activities were progressively inhibited with increasing numbers of phosphorothioate modifications in the crRNA spacer region, indicating that the CasTAC strategy enables tunable down-modulation of Cas12a activity in a dose-dependent manner (Fig. 3E and F; Supplementary Figs S24 and S25). Furthermore, pscrRNA20 produced a markedly stronger fluorescence signal in the one-pot RPA−CRISPR/LbCas12a system (Fig. 3G and Supplementary Fig. S26). This enhancement can be attributed to the slower kinetics of the CasTAC system, which enables more amplification products to accumulate for subsequent LbCas12a activation, thereby improving the fluorescence signal. Moreover, experimental results demonstrated that the introduction of pscrRNA did not adversely affect RPA amplification efficiency (Supplementary Fig. S27). This observation may be attributed to the abundant and diverse protein components within the RPA system, which likely mitigate potential inhibitory effects of phosphorothioate regulators.

We further explored the molecular mechanism of the inhibitory effect of the CasTAC strategy. Molecular docking results revealed that pscrRNA extensively interacts with multiple functional domains of LbCas12a, including the recognition domains (WED-I, REC1, and REC2) and core catalytic regions (WED-III, RuvC-I, RuvC-II, RuvC-III, and Nuc) (Fig. 3H). Notably, the overall binding mode of pscrRNA was highly similar to that of canonical crRNA, indicating that phosphorothioate regulators within the spacer region do not substantially disrupt the native crRNA-LbCas12a recognition pattern. Meanwhile, the phosphorothioate-modified spacer region was predicted to localize near the recognition and catalytic channels, which may contribute to inhibition of substrate access and catalytic activity. In addition, binding energy calculations showed that the binding energy of pscrRNA to LbCas12a (−408.39 kJ·mol⁻¹) was considerably lower than that of crRNA (−150.40 kJ·mol⁻¹), suggesting enhanced binding affinity following phosphorothioate modification (Fig. 3I). This result was further supported by EMSA analysis, which indicated stronger binding between pscrRNA and LbCas12a compared with unmodified crRNA (Fig. 3J). Together, these findings provide mechanistic support for the enhanced interaction between pscrRNA and LbCas12a.

The stability of crRNA during storage and application is an important factor influencing the performance of CRISPR-based diagnostics, and engineering crRNA to improve robustness is critical for practical deployment. Previously, we demonstrated that circularization of crRNA could protect terminal ends from nuclease degradation, yet the complexity of preparation and low yield limited its utility [34]. Phosphorothioate-anchored nucleic acids are known to exhibit enhanced nuclease resistance compared with unmodified counterparts [35]. To assess whether this modification could improve crRNA stability, we subjected both unmodified crRNA and pscrRNA to RNase R digestion for varying durations and analyzed the products using PAGE and CRISPR trans-cleavage assays (Supplementary Fig. S28A). Whereas unmodified crRNA was almost completely degraded within 5 min, pscrRNA remained detectable for up to 1 h (Supplementary Fig. S28B and C). CRISPR trans-cleavage assays further supported the improved nuclease resistance of pscrRNA (Supplementary Fig. S28D and E). Furthermore, we introduced RNase R into the CRISPR system to simulate ribonuclease contamination (Supplementary Fig. S29A). The trans-cleavage activity of the conventional CRISPR system decreased markedly with increasing RNase R concentrations (Supplementary Fig. S29B), while that of the CasTAC system maintained above 80% (Supplementary Fig. S29C). Collectively, these findings demonstrate that the CasTAC system provides improved stability compared with the conventional CRISPR system, underscoring its potential for practical diagnostic applications.

Single-nucleotide specificity of CasTAC system

The identification of single-nucleotide variants requires high specificity of the CRISPR/Cas12a system. However, conventional CRISPR systems have been reported to exhibit insufficient specificity to date [36]. To evaluate the single-nucleotide discrimination capability, single-nucleotide mutations were introduced at multiple sites within the target-binding region of three distinct targets (Fig. 4A−C). We then systematically investigated how the introduction of phosphorothioate regulators in the crRNA spacer modulates the specificity of the CRISPR/Cas12a system, and further quantified performance using the discrimination factor (DF). We observed that phosphorothioate modification significantly enhanced single-nucleotide discrimination at multiple sites (Fig. 4D−F). Notably, specificity was markedly improved within the seed region of crRNA, and this finding was validated across three targets with completely different sequences. Collectively, these results suggest that our proposed CasTAC system possesses improved specificity compared with the conventional CRISPR/Cas12a system.

Figure 4.

Enhanced single-nucleotide specificity of CasTAC system. Sequence information of HPV-16 (A), H1N1 (B), and MTB (C) templates with single-nucleotide mutation. Compare the DF values of the conventional CRISPR-Cas12a system and the CasTAC system for HPV-16 (D), H1N1 (E), and MTB (F) detection. (G) Schematic diagram illustrating the mechanism of specificity enhancement of the CasTAC system. (H) Molecular docking simulations of the LbCas12a/crRNA_HPV or LbCas12a/pscrRNA20_HPV interacting with off-target dsDNA (dsDNA_HPV-MT9). (I) Comparison of binding energies between LbCas12a/crRNA/dsDNA_HPV-MT9 and LbCas12a/pscrRNA/dsDNA_HPV-MT9 complexes. (J) Calculation relative distance from the off-target dsDNA to the LbCas12a/crRNA_HPV and LbCas12a/pscrRNA20_HPV. The concentration of all dsDNA used in the exper with DF formula: DF = (F_WT – background)/(F_MT – background). Data represent mean ± SD (n = 3 replicates).

Previous studies have shown that phosphorothioate modification in double-stranded nucleic acids reduces the stability of the nucleic acid duplex [37]. Accordingly, we hypothesized that introducing phosphorothioate modifications into pscrRNA would weaken the stability of the ternary complex assembled by the LbCas12a/pscrRNA complex and the target dsDNA. Currently, it has been well documented that moderate destabilization of the ternary complex improves CRISPR specificity [38, 39, 40]. Based on this, we speculate that the high specificity of the CasTAC system also results from the reduced stability of the ternary complex (Fig. 4G). We first employed molecular docking to assess the stability of the ternary complex composed of LbCas12a/pscrRNA and off-target dsDNA (Fig. 4H). The calculated binding energies were –881.52 kJ·mol⁻¹ for the LbCas12a/crRNA/dsDNA_HPV-MT9 complex and –831.16 kJ·mol⁻¹ for the LbCas12a/pscrRNA/dsDNA_HPV-MT9 complex (Fig. 4I). A higher binding energy corresponds to weaker intermolecular binding affinity. This result implies that phosphorothioate modification reduces the binding ability of the LbCas12a/pscrRNA complex to off-target targets, thereby decreasing the structural stability of the ternary complex. Furthermore, in the LbCas12a/pscrRNA/dsDNA_HPV-MT9 complex, the relative distance between the off-target dsDNA and the Cas12a surface increased by 3.9–46.4 Å. This structural difference further supports that phosphorothioate modification destabilizes the stability of the ternary complex at the spatial conformational level (Fig. 4J).

In addition, reduced ternary complex stability also could decelerate the hybridization kinetics between crRNA and target dsDNA [2, 41]. To evaluate the effect of phosphorothioate modification on the formation of the Cas12a/crRNA/target ternary complex, we designed a molecular beacon-based assay. The assembled beacon is composed of three oligonucleotide fragments: a 37-nt TS with 5′ end FAM fluorophore (Beacon-FAM), a 20-nt non-TS with 3′ end BHQ quencher (Beacon-BHQ), and a 17-nt sequence containing the TTTA PAM motif (Beacon-17 nt). These components anneal to form a nicked dsDNA. Due to energy transfer via FRET between the fluorescent and quencher groups, the initial fluorescence intensity of the beacon is low. Upon binding of Cas12a-crRNA complex to the beacon substrate, duplex unwinding and strand separation occur to physically isolate the fluorophore from its quencher, which markedly elevates the detectable fluorescence readout. As shown in Supplementary Fig. S30, the hybridization rate mediated by pscrRNA20 was markedly lower than that of wild-type crRNA, directly proving that the phosphorothioate modification was destabilizing the the Cas12a/crRNA/target ternary complex. Collectively, these results indicate that incorporation of phosphorothioate regulators into crRNA destabilizes the ternary complex and consequently enhances the single-base recognition fidelity of the CasTAC system. In summary, the proposed CasTAC system enables flexible modulation of activity tailored to specific applications while significantly improving specificity compared with conventional CRISPR systems. These advances are expected to mark an advance in targeting efficiency and accuracy of molecular diagnostics and gene editing.

One-pot, one-step RPA−CasTAC assay

Recent viral outbreaks have highlighted the need for rapid and ultrasensitive nucleic acid detection technologies [42, 43]. CRISPR-based diagnostics provide a simple and field-deployable alternative to qPCR. However, early CRISPR diagnostic platforms generally required separate nucleic acid amplification and CRISPR cleavage steps, increasing workflow complexity and the risk of aerosol contamination. Integrating nucleic acid amplification and CRISPR/Cas12a into a single-tube reaction has therefore emerged as an attractive strategy for simplifying operation and minimizing contamination risk. Nevertheless, the major challenge of one-pot assays arises from kinetic interference between amplification and premature CRISPR activation. During the early stage of amplification, trace amounts of amplicons can rapidly activate Cas12a cleavage activity before sufficient amplification products accumulate. Premature Cas12a activation may interfere with the amplification process through cleavage of amplification templates or primers, thereby suppressing amplification efficiency and reducing overall detection sensitivity. Because efficient amplification relies on an exponential amplification phase, balancing amplification kinetics and CRISPR cleavage activity remains a critical challenge for highly sensitive one-pot assays. To overcome this barrier, early efforts mainly focused on optimizing reaction components to balance the kinetics of RPA and CRISPR cleavage within a one-pot systems [44, 45]. Alternative approaches integrated both workflows into a closed system via physical isolation strategies, including using simple spatial isolation, aqueous multiphase reaction system, and microfluidic chips [46–48]. However, these approaches increase design complexity, require extra handling steps, and reduce the reproducibility of results. Improved strategies utilize photosensitive molecules to temporarily inhibit the activity of Cas12a until reactivation by light [13–15, 49, 50]. Although delayed activation by light irradiation is more flexible than physical isolation strategies, it still relies on external stimulation and therefore cannot achieve a fully autonomous one-step reaction. More recent approaches attempted to reduce CRISPR cleavage kinetics directly, for example by screening suboptimal PAM sequences or engineering the direct repeat region of crRNA to weaken Cas12a activity [51, 52]. Although these strategies partially improved one-pot analytical performance, they generally require laborious optimization and often suffer from limited generalizability across different targets and systems.

To achieve selective and tunable regulation of Cas12a activity, we proposed a CasTAC strategy. In this approach, crRNA serves as proximity mediator to carry phosphorothioate regulators in close proximity to Cas12a, thereby achieving specific modulation of its activity in multi-enzyme systems. This regulatory mechanism is expected to resolve compatibility issues in one-pot assays, allowing efficient accumulation of amplification products prior to Cas12a activation. Based on this principle, we developed a one-pot, one-step RPA−CasTAC assay (Fig. 5A). To evaluate the sensitivity of the RPA−CasTAC system, the HPV-16 gene was selected as a representative target. Using gradient-diluted synthetic HPV-16 plasmids, the RPA−CasTAC system achieved a 1000-fold improvement in detection sensitivity compared with the conventional one-pot RPA–CRISPR/Cas12a assay (Fig. 5B and C).

Figure 5.

Development of the one-pot, one-step RPA-CasTAC assay. (A) Schematic diagram of the conventional one-pot RPA-CRISPR/Cas12a assay and the one-pot, one-step RPA-CasTAC assay. Detection sensitivity of conventional one-pot RPA-CRISPR/Cas12a assay (B) and the one-pot, one-step RPA-CasTAC assay (C) for detecting serially diluted HPV-16 plasmid. The right panel is real-time fluorescence curves, and the histogram in the left panel represents RFU at 90 min. The real-time fluorescence analysis (D) and agarose gel analysis (E) of the cleavage of RPA amplification products by CRISPR/Cas12a and CasTAC in a one-pot system. Peak time was obtained by monitoring the real-time fluorescence of RPA amplification using SYBR Green as the signal reporter. The target HPV-16 plasmid concentrations for real-time fluorescence analysis and agarose gel analysis were 10 fg/μl and 100 fg/μl, respectively. RPA refers to the reaction with only RPA amplification. RPA-crRNA refers to the one-pot RPA-CRISPR/Cas12a assay conducted using crRNA_HPV. RPA-pscrRNA refers to the one-pot, one-step RPA-CasTAC assay conducted using pscrRNA20_HPV. Detection sensitivity of conventional one-pot RPA-CRISPR/Cas12a assay (F) and the one-pot, one-step RPA-CasTAC assay (G) for detecting serially diluted H1N1 plasmid. The right panel is real time fluorescence change curves, and the histogram in the left panel represents RFU at 90 min. Data represent mean ± SD (n = 3 replicates). Two-tailed Student’s t-test; *P < .1; **P < .01; ***P < .001; ^****P < .0001; ns, not significant.

Effective detection in one-pot assays depends on a delicate balance between RPA amplification and CRISPR cleavage, as both processes compete for the same DNA templates. Excessive cis-cleavage by Cas12a prematurely depletes templates and compromises amplification, whereas attenuated cleavage preserves sufficient templates for efficient RPA, thereby enhancing detectability. To elucidate this balance, we further examined the nucleic acid amplification within the one-pot system. Real-time fluorescence results revealed that CRISPR cleavage markedly suppressed RPA amplification in conventional RPA−CRISPR system, whereas the RPA−CasTAC system maintained a fluorescence intensity comparable to the crRNA-free control (Fig. 5D and Supplementary Fig. S31). Consistently, PAGE analysis showed rapid accumulation of amplification products within 10 min in the RPA–CasTAC system, in sharp contrast to negligible amplification in the conventional RPA–CRISPR system (Fig. 5E). Collectively, these findings indicate that the RPA–CasTAC system enables more effective coordination of RPA and CRISPR kinetics, thereby accounting for its substantially enhanced detection sensitivity. To further demonstrate the universality of the CasTAC system, we evaluated its detection sensitivity for the H1N1 gene. The CasTAC system for H1N1 plasmid detection achieves 10 000-fold higher detection sensitivity than the conventional one-pot RPA–CRISPR/Cas12a assay (Fig. 5F and G).

Clinical applicability of the RPA−CasTAC assay

Tuberculosis, caused by MTB, remains a major global health threat, highlighting the urgent demand for rapid, sensitive, and field-deployable diagnostic technologies [53]. Following validation of the sensitivity and universality of the one-pot RPA–CasTAC assay, we next assessed its clinical applicability using MTB as a model target (Fig. 6A). We selected the MTB-specific insertion sequence IS6110 as the detection target, and optimized key reaction parameters, including Mg²⁺ and primer concentrations (Supplementary Fig. S32). Under optimal conditions, the one-pot RPA-CasTAC assay achieved a detection limit of 10 aM for synthesized plasmids, representing a 10 000-fold improvement in sensitivity compared with conventional one-pot RPA–CRISPR/Cas12a assay (Supplementary Fig. S33). To evaluate clinical performance, we analyzed 80 bronchoalveolar lavage fluid (BALF) samples collected from suspected tuberculosis patients. The samples were pre-classified by qPCR, which identified 59 positive and 21 negative cases (Fig. 6B, Supplementary Figs S34–S37, and Supplementary Table S3). The one-pot RPA–CasTAC assay correctly detected 54 of the qPCR-positive samples and all negative samples, corresponding to a sensitivity of 91.5% and specificity of 100% (Fig. 6C–E, Supplementary Fig. S38 and S39). Receiver operating characteristic (ROC) analysis yielded an area under the curve (AUC) value of 0.9516 (Fig. 6F), underscoring the diagnostic potential of the one-pot RPA–CasTAC assay. To further evaluate robustness across different sample matrices, we additionally incorporated 25 sputum samples for parallel analysis. The samples were also pre-classified by qPCR, which identified 16 positive and 9 negative cases (Fig. 6G, Supplementary Figs S40 and S41, and Supplementary Table S3). Our assay correctly identified 15 of the qPCR-positive samples and 8 negative samples, corresponding to a sensitivity of 93.8% and specificity of 88.9% (Fig. 6H–J, Supplementary Fig. S42 and S43). ROC analysis yielded an AUC value of 0.9653 (Fig. 6K). The effective detection performance of the one-pot RPA–CasTAC assay across two distinct sample types validates the clinical application potential of this strategy.

Figure 6.

Clinical evaluation of the one-pot, one-step RPA-CasTAC system for MTB detection. (A) Workflow for analyzing MTB-infected clinical samples using the one-pot RPA-CasTAC assay. Heat map displaying the detection results of 80 BALF samples measured by qPCR (B) and RPA–CasTAC (C). (D) Scatter plot differentiating MTB-positive from MTB-negative samples based on fluorescence signal. (E) Concordance table between one-pot RPA-CasTAC assay and qPCR for the 80 BALF samples. (F) ROC curve evaluating the clinical detection accuracy of the one-pot RPA-CasTAC assay for BALF samples. Heat map displaying the detection results of 25 sputum samples measured by qPCR (G) and RPA–CasTAC (H). (I) Scatter plot differentiating MTB-positive from MTB-negative samples based on fluorescence signal. (J) Concordance table between one-pot RPA-CasTAC assay and qPCR for the 25 sputum samples. (K) ROC curve evaluating the clinical detection accuracy of the one-pot RPA-CasTAC assay for sputum sample. Data represent mean ± SD (n = 3 replicates).

Discussion

In summary, we present a CasTAC strategy to eliminate nonspecific regulation and improve tunability, thereby enabling selective and tunable regulation of Cas12a activity. In this approach, crRNA functions as a proximity mediator to carry phosphorothioate regulators near Cas12a, allowing fine-tuned modulation of its catalytic behavior in multi-enzyme systems. Despite a recent report describing psDNA-mediated inhibition of Cas12a cleavage activity during the course of our manuscript preparation, this system relies on freely diffusing photocleavable phosphorothioate-modified DNAs for Cas12a modulation [17]. As noted in that published literature, free phosphorothioate nucleic acids readily produce undesired nonspecific binding with diverse proteins, which severely restricts their practical deployment within complex multi-enzyme reaction environments. Differing fundamentally from such free-regulator-based regulatory modes, our CasTAC approach achieves proximity-guided and specific regulation of Cas12a activity while minimizing nonspecific interactions of free phosphorothioate regulators in complex molecular systems, serving as the core superiority for accomplishing reliable selective regulation of LbCas12a within multi-enzyme systems.

We systematically investigated the effect of phosphorothioate-anchored nucleic acids on the DNase (cis- and trans-cleavage) and RNase activity of Cas12a. Biochemical assays and molecular docking revealed that phosphorothioate-anchored nucleic acids possess markedly higher binding affinity for LbCas12a, following the order pscrRNA > psDNA > crRNA. These phosphorothioate regulators interfere with both catalytic and recognition domains through high-affinity interactions, thereby suppressing LbCas12a activity. Through rational phosphorothioate engineering of crRNA, LbCas12a activity can be selectively tuned within the RPA system composed of recombinase-loading factors, recombinase, creatine kinase, DNA polymerase, and single-stranded binding proteins. Moreover, LbCas12a activity can be finely adjusted to different inhibitory levels by varying the number of phosphorothioate regulators, realizing programmable inhibitory regulation of Cas12a activity. Given the mature chemistry of phosphorothioate modification, the CasTAC strategy provides a general and flexible means to achieve programmable protein regulation.

In contrast to previous regulators (e.g. heparin sodium or phosphorothioate-modified DNA) that act via nonspecific electrostatic interactions, the CasTAC system achieves sequence- and structure-dependent recognition, thereby eliminating nonspecific binding and reaction interference. The CasTAC system also exhibits enhanced nuclease resistance and superior single-nucleotide discrimination compared with conventional CRISPR/Cas system, which is expected to mark an advance in the sensitivity of nucleic acid detection and the accuracy of gene editing. Furthermore, the CasTAC system effectively balances the kinetics of nucleic acid amplification and CRISPR cleavage, facilitating efficient product accumulation and resolving compatibility issues in one-pot assays. Based on this delayed activation, we developed a one-pot, one-step RPA–CasTAC assay that achieved over 1000-fold higher detection sensitivity than the conventional one-pot RPA–CRISPR/Cas12a assay. In clinical evaluation with 105 suspected tuberculosis samples, the RPA–CasTAC assay achieved a sensitivity of 91.5% and a specificity of 100% for BALF samples and 93.8% sensitivity alongside a sensitivity of 93.8% and a specificity of 88.9%, consistent with qPCR results. Importantly, this CasTAC strategy requires no external stimuli, offering operational simplicity and robust adaptability.

Although the CasTAC strategy was successfully validated in this study, several important limitations and future challenges should be acknowledged. First, the current mechanistic understanding of CasTAC remains primarily based on molecular docking, biochemical binding analysis, and functional characterization. Although these results collectively support the proposed proximity-guided regulatory mechanism, they do not yet provide direct high-resolution structural evidence regarding how phosphorothioate regulators influence the conformational dynamics and catalytic accessibility of Cas12a. Future structural studies, including cryo-electron microscopy and long-timescale molecular dynamics simulations, will strengthen the mechanistic understanding of the CasTAC strategy.

Second, the present work mainly emphasizes methodological development and preliminary translational validation for in vitro molecular diagnostics. Although favorable performance was achieved in clinical samples, the current validation cohort remains relatively limited in size and diversity. Larger-scale prospective clinical studies involving different disease stages and complex co-infection scenarios are of great significance for further assessing the practical applicability of the CasTAC-based detection platform.

Finally, the current CasTAC system mainly enables selective and tunable control of Cas12a activity, whereas reversible regulation has not yet been explored. Incorporation of stimulus-responsive regulatory modules may further expand the versatility of this strategy in complex biological systems. In addition, because the present study was restricted to in vitro diagnostic applications, comprehensive evaluation of long-term biosafety, immunogenicity, pharmacokinetics, and metabolic effects was beyond the scope of this work but will be essential for future in vivo therapeutic or gene-regulation applications.

Overall, the CasTAC strategy advances CRISPR technology toward refined, context-adaptable functionality and is expected to broaden its potential in genome engineering and next-generation molecular diagnostics. Central to this approach is the use of crRNA as a high-affinity module to carry phosphorothioate regulators to regulate the activity of Cas12a. CasTAC establishes a paradigm for targeted protein modulation by integrating universal activity regulators with an induced proximity mechanism. Although our current study focuses on Cas12a, the strategy could be extended to regulate other proteins. By selecting target-specific and high-affinity nucleic acid strands as mediators to carry phosphorothioate regulators, this approach enables the assembly of targeting chimeras. It may thereby evolve into a general framework for selective protein activity control, applicable to the targeted modulation of specific proteins across diverse molecular networks.

Supplementary data

Supplementary data is available at NAR online.

Conflict of interest

None declared.

Funding

This work is supported by the National Natural Science Foundation of China (82072378, 22504009), Natural Science Foundation of Chongqing, China (CSTB2025NSCQ-LZX0126, CSTB2025NSCQ-GPX0375), Chongqing Talents: Exceptional Young Talents Project, China (CQYC202005015, cstc2021ycjh-bgzxm0328). Funding to pay the Open Access publication charges for this article was provided by the research grants supporting this study.

Data availability

The data supporting this article are available through Figshare at https://doi.org/10.6084/m9.figshare.33025166.