Programmable hooded DNA switches for conditional control of CRISPR/Cas12a in multiplexed biosensing

Abstract

The CRISPR/Cas system has become an indispensable tool for programmable and accurate biosensing, with its performance critically dependent on precise activity control. While most regulatory strategies have focused on engineering Cas proteins or modifying CRISPR RNAs, relatively little attention has been given to the design of substrate probes. Here, we systematically characterize the trans-cleavage activity of split CRISPR/Cas12a on structured substrates and leverage this insight to engineer a tunable “Hooded” probe with switchable properties. This probe architecture confers protection against trans-cleavage, and its activity can be progressively modulated by varying the probe length. Utilizing this design, we constructed a multiplexed logic-gated detection platform for direct and simultaneous analysis of miRNA and PSA, which demonstrated high sensitivity and specificity. Furthermore, we validated the robust performance of this system for logic-operated imaging in diverse cellular models, confirming its reliability in complex biological settings. Overall, our Hooded probe strategy not only broadens the applicability of CRISPR/Cas12a in molecular diagnostics, but also provides a novel design principle for the multiplexed biosensing.

Graphical abstract

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-026-04122-w.

Introduction

The CRISPR/Cas system represents a groundbreaking advance in genetic engineering and has become an indispensable tool for gene editing and biosensing applications [1, 2]. Owing to its exceptional programmability, high substrate specificity, and efficient signal amplification, CRISPR/Cas has emerged as a powerful platform for in vitro diagnostics and bioimaging [3–5]. Furthermore, in the era of precision medicine, achieving precise spatiotemporal control over CRISPR/Cas activity is increasingly recognized as pivotal, both for enhancing detection reliability and for unlocking its full therapeutic potential [6].

The CRISPR/Cas system employed in biosensing comprises three essential components: the Cas protein, CRISPR RNA (crRNA), and a reporter probe that primarily relies on Cas’s trans-cleavage activity. Numerous studies have previously investigated the reaction characteristics and regulatory strategies of the CRISPR/Cas system by focusing on the first two components. The Cas protein serves as the core of the system, responsible for cleavage activity. Cas proteins derived from various bacterial species exhibit distinct enzymatic activities, among which widely used variants include Cas9 [7], Cas12a [8], and Cas13a [9]. Protein engineering of Cas proteins allows modification of their reaction conditions, cleavage activity, or introduction of stimulus-responsive properties [10, 11]. However, due to the challenges and high cost associated with protein modification, increasing attention has been directed toward crRNA. crRNA consists of one or more RNA sequences responsible for target recognition. Modulations such as sequence adjustments involving structure, length, and content [12, 13], structural alterations [14, 15], or chemical modifications [16, 17] can enhance or diminish its targeting capacity and substrate discrimination, enabling precise control over catalytic activity. Recently, several studies have reported that split crRNAs, consisting of a scaffold RNA and a spacer RNA, retain the ability to activate Cas12a comparable to intact crRNA [18–20]. This innovation extends target detection from DNA to RNA, significantly improving versatility and ease of use. In contrast to crRNA-directed strategies, modulation through the reporter probe also substantially influences biosensing performance [21, 22]. Furthermore, reporter probes function as relatively independent components within the system and do not interfere with the fundamental properties of the CRISPR/Cas system. Their structural and chemical flexibility exceeds that of crRNA, thereby conferring greater biosafety and programmability. Nevertheless, although several studies have examined the affinity and cleavage efficiency of Cas proteins toward different substrate probes, they have rarely been utilized as regulatory elements in CRISPR/Cas-mediated biosensing [23]. Therefore, it is imperative to develop novel and flexible probe tools to diversify the biosensing strategies and application pathways of CRISPR/Cas-based technologies.

Herein, we investigated the cleavage activity of the CRISPR/Cas12a system toward double-stranded substrates with varying structures and demonstrated that Hooded (HAT-shaped) structures confer strong resistance to trans-cleavage (Scheme 1). Furthermore, exploration of allosteric DNA probes revealed that the trans-cleavage activity of the CRISPR/Cas system depends critically on the presence of Hooded structures. By integrating the switch-like behavior of Hooded probes with a split CRISPR/Cas12a system, we developed a multiplexed logic detection system. Based on this design, we constructed a simple and highly specific liquid biopsy platform for prostate cancer (PCa) detection. Moreover, we applied the logic system for imaging in multiple live cell types, confirming its reliability and broad applicability across diverse biological environments. By integrating the switch-like behavior of Hooded probes with a split CRISPR/Cas12a system, we developed a multiplexed logic detection system. based on it, we constructed a straightforward and highly specific liquid biopsy platform for PCa diagnostic. Furthermore, we implemented the logic system for live-cell imaging across multiple cell types, demonstrating its reliability and applicability in diverse biological environments.

Scheme 1.

Regulation of trans-cleavage activity in split CRISPR/Cas12a using Hooded DNA probes. (A) Compared to conventional double-stranded DNA (dsDNA), Hooded DNA substrates exhibit enhanced resistance to trans-cleavage. (B) The trans-cleavage activity of split CRISPR/Cas12a can be continuously tuned by programming the size of the Hooded DNA structure. (C) Hooded DNA serves as a molecular signal-controlled switch, enabling applications in multidimensional in vitro diagnostics and intracellular imaging

Materials and methods

Materials

All oligonucleotides used in this study were synthesized and purified by Sangon Biotech Co., Ltd. (Shanghai, China). The concentrations of DNA oligonucleotides were measured using a NanoDrop 2000 UV-Vis spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). The sequences of all oligonucleotides are listed in Table S1-S6. AsCas12a, FnCas12a, and LbCas12a were purchased from Tolo Biotech Co., Ltd. (Shanghai, China). MgSO₄ (100 mM) was obtained from New England Biolabs Inc. (Beijing, China). Agarose, 5× TBE buffer, 4S Red Plus Nucleic Acid Stain, 6× DNA loading dye, DNA marker, and ATP were acquired from Sangon Biotech Co., Ltd. (Shanghai, China). The MiPure Cell/Tissue miRNA Kit, miRNA 1st Strand cDNA Synthesis Kit (by stem-loop), and miRNA Unimodal SYBR qPCR Master Mix were purchased from Vazyme (Jiangsu, China). The Human IgG (HY-NP189) was acquired from MedChemExpress (MCE, Shanghai, China).

Cell lines and cell culture

A total of six cell lines were used in this study, all of which were obtained from BOSTER (Wuhan, China). Three PCa cell lines (22Rv1, C4-2, and LNCaP), one bladder cancer cell line (5637), and the commonly used tool cell line 293T were cultured in RPMI-1640 medium supplemented with penicillin G (100 U/mL), streptomycin (100 µg/mL), and 10% fetal bovine serum (FBS). The human normal prostate epithelial cell line RWPE-1 was maintained in a specialized keratinocyte serum-free medium (K-SFM) (IMMOCELL, Xiamen, China) containing 0.05 mg/ml bovine pituitary extract (BPE), 5 ng/ml epidermal growth factor (EGF), and penicillin-streptomycin (PS). All cells were incubated at 39 °C in a humidified atmosphere with 5% CO₂.

Collection and processing of plasma samples

Fresh plasma samples were collected from 48 patients with benign prostatic hyperplasia (BPH) and 48 patients with PCa at Tongji Hospital in Wuhan. miRNA and proteins were extracted from the plasma using the miRNeasy Serum/Plasma Advanced Kit (QIAGEN, Germany) and the Ultra Plasma Protein Extraction Kit (Solarbio, China), respectively, and stored at -80 °C. The clinical study protocol was approved by the Ethics Committee of Tongji Hospital (Approval No: TJ-IRB202407023), and informed consent was obtained from all participants.

Fluorescence signal acquisition and normalization

Fluorescence signals were recorded every 20 s using a qPCR instrument (Bioer, China). The excitation and emission wavelengths were set at 470 nm and 510 nm for the FAM channel. ΔFluorescence intensity was calculated as the rate of fluorescence increase per minute.

Electrophoresis

Agarose gel electrophoresis was performed using 3% agarose gels pre-stained with 4S Red Plus nucleic acid stain in 1× TBE buffer at 110 V. A 10 µL aliquot of each sample was mixed with 2 µL of 6× DNA loading buffer before loading into the wells. The gels were visualized using a Bio-Rad Universal Hood II gel imaging system (Bio-Rad, Shanghai, China).

Direct detection of MiRNA using split CRISPR/Cas12a

To a 200 µL PCR tube, sequentially add 2 µL scaffold RNA (5 μM), 2 µL spacer RNA (5 μM), 2 µL DNA (5 μM), 2 µL probe (5 μM), 4 µL of Mg²⁺ (100 mM), 1 µL LbCas12a (10 μM), 5 µL buffer r2.1 (10×) and brought up to a total volume of 50 µL by DEPC water. Incubate the reaction at 39 °C in a PCR machine while monitoring the fluorescence intensity.

Dependence of CRISPR/Cas12a trans-cleavage activity on the hooded probe

To form a stable HAT Probe, 5 µM each of the HAT and HAT-probe DNA single strands were mixed in 1× buffer r2.1. The mixture was heated at 85 °C for 5 min and then annealed at 45 °C for 10 min. Then, to a 200 µL PCR tube, sequentially add 2 µL scaffold RNA (5 μM), 2 µL spacer RNA (5 μM), 2 µL DNA (5 μM), 2uL HAT Probe (5 μM), 4 µL Mg²⁺ (100 mM), 1 µL LbCas12a (10 μM), 5 µL buffer r2.1 (10×) and brought up to a total volume of 50 µL by DEPC water. Incubate the reaction at 39 °C in a PCR machine while monitoring the fluorescence intensity in real time.

DNAzyme-based detection of PSA

To form a stable DNAzyme-PSA-Aptamer (DPA), 5 µM DPA was prepared in 1× buffer r2.1. The solution was heated at 85 °C for 5 min and annealed at 45 °C for 10 min. Then, to a 200 µL PCR tube, sequentially add 2uL DPA (5 μM), varying concentrations of PSA, 5 µL buffer r2.1 (10×) and brought up to a total volume of 48 µL by DEPC water. Then incubate at 37 °C for 1 h. Subsequently, add 2 µL probe (5 μM) and incubate the reaction at 39 °C in a PCR machine while monitoring the fluorescence intensity in real time.

Multiplex detection of MiRNA and PSA

To a 200 µL PCR tube, add 2uL DPA (5 μM), varying concentrations of PSA, 4 µL buffer r2.1 (10×) and brought up to a total volume of 37 µL by DEPC water. Then incubate at 37 °C for 1 h. Then add 5uL HAT-Probe@AuNP (2 μM) and continue incubation at 37 °C for 1 h. Next, sequentially add 2uL scaffold RNA (5 μM), 2uL spacer RNA (5 μM), 2uL DNA (5 μM), 4uL Mg²⁺ (100 mM), 1 µL buffer r2.1 (10×), 1 uL LbCas12a protein (10 μM), and incubate at 39 °C while monitoring the fluorescence intensity in real time.

Construction of the HAT-functionalized AuNPs (HAT-Probe@AuNP)

AuNPs with a diameter of 20 nm were mixed with FAM-labeled HAT Probe at a molar ratio of 1:1000, following a previously described protocol [24]. The resulting HAT-Probe@AuNP were stored in 25 mM Tris-HC1 (pH 7.4) at a concentration of 2 nM and kept at 4 °C until use.

Evaluation of cell death ratio

22RV1 cells were seeded in 6-well plates at a density of approximately 1 × 10⁶ cells per well. After necessary experimental treatments, the cells were incubated for 24 h at 37 °C with 5% CO₂. Subsequently, the cells were stained with Calcein-AM and propidium iodide (PI) (Beyotime Biotech, Shanghai, China) for 10 min. Calcein-AM labels live cells with green fluorescence, whereas PI labels dead cells with red fluorescence. Cell imaging was performed using a fluorescence microscope (Bio-Rad, USA). The fluorescence intensity of the images was quantified using FiJi software (https://imagej.net/software/fiji/).

Electroporation procedure

Electroporation was performed using an ECM 830 square-wave electroporation system (BTX, Harvard Apparatus). Cells were harvested and resuspended in phosphate-buffered saline (PBS) to obtain a uniform cell suspension. The cell suspension was mixed with the indicated probes or constructs and transferred to electroporation cuvettes with a fixed gap width of 4 mm. Electroporation was carried out using a single square-wave pulse at a voltage of 230 V with a pulse duration of 4 ms. No additional pulses or auxiliary electrical parameters were applied. Following electroporation, cells were immediately transferred to pre-warmed complete culture medium and incubated under standard cell culture conditions prior to subsequent analysis. All electroporation experiments were performed using identical electrical and buffer conditions to ensure reproducibility.

Structural models and molecular docking

Molecular simulations were based on the X-ray crystal structure of the CRISPR/Cas12a system from Francisella novicida in complex with crRNA and target DNA (PDB ID: 5NFV). This structure was used as the starting model for all simulations. The designed HAT and Hairpin DNA reporters were modelled to generate initial three-dimensional conformations using AlphaFold3. The resulting DNA structures were further refined and geometry-optimized in PyMOL using the Optimize plugin with the General Amber Force Field (GAFF). The optimized DNA reporters were subsequently assembled with the Cas12a-crRNA complex to generate the corresponding protein-DNA complexes used for docking and molecular dynamics simulations. Molecular docking of the HAT and Hairpin DNA reporters to Cas12a was performed using HADDOCK 2.4 (https://rascar.science.uu.nl/haddock2.4/). The catalytic residues D917, E1020, and D1255 were defined as active residues, while all other parameters were kept at their default settings. The most representative docked conformations, as determined by HADDOCK scoring and clustering, were selected as starting structures for subsequent molecular dynamics simulations.

Molecular dynamics simulations

All-atom molecular dynamics simulations were carried out using GROMACS (version 2023.2). The AMBER99SB-ILDN force field was employed for the protein, while nucleic acids were described using compatible AMBER parameters. Each protein-DNA complex was solvated in explicit TIP3P water molecules in a periodic cubic simulation box, with a minimum distance of 1.2 nm between the solute and the box boundary. Appropriate numbers of Na⁺ and Cl⁻ ions were added to neutralize the systems and to mimic physiological ionic conditions. Energy minimization was first performed using the steepest descent algorithm. The systems were then equilibrated under NPT conditions, followed by 100 ns production MD simulations for each Cas12a-DNA complex. All simulations were performed with a time step of 2 fs, and covalent bonds involving hydrogen atoms were constrained using the LINCS algorithm. Long-range electrostatic interactions were treated using the particle mesh Ewald (PME) method with a real-space cutoff of 1.2 nm. Van der Waals interactions were truncated at 1.2 nm using the Verlet cutoff scheme. Temperature was maintained at 310 K using the velocity-rescale (V-rescale) thermostat with a coupling constant of 0.1 ps. Pressure was controlled at 1 bar using the Parrinello-Rahman barostat with isotropic coupling and a time constant of 2.0 ps. Periodic boundary conditions were applied in all three spatial dimensions throughout the simulations. Due to limitations in metal ion parameterization during the docking process, Mg²⁺ ions were removed prior to docking and were not included in the MD simulations. This treatment was consistently applied across all systems to ensure methodological consistency and comparability.

Binding free energy calculations and trajectory analysis

Binding free energies of the Cas12a-DNA complexes were estimated using the molecular mechanics generalized Born surface area (MM-GBSA) method as implemented in the gmx_MMPBSA package. MM-GBSA calculations were performed based on the last 50 ns of each MD trajectory, during which the systems exhibited stable conformational behavior. Structural stability and protein-DNA interactions were further analyzed using root mean square deviation (RMSD), minimum distance, contact number, and hydrogen bond analyses. All analyses were conducted using the same stable trajectory window to ensure consistency between structural and energetic evaluations.

Statistical analysis

Data are presented as mean ± standard deviation (SD). Statistical differences between groups were analyzed using one-way analysis of variance (ANOVA) followed by Tukey’s post hoc test in GraphPad Prism (version 9.5). A two-sided P value < 0.05 was considered statistically significant. Differential expression and feature selection analyses of miRNAs were performed using publicly available datasets from The Cancer Genome Atlas (TCGA). Data processing and analysis were conducted in R (version 4.4.1) with machine learning-based feature selection approaches, including least absolute shrinkage and selection operator (LASSO) regression, support vector machine (SVM)-based classification, and random forest (RF) analysis. Figures were created using BioRender (https://www.biorender.com/), for which publication permission has been obtained.

Results and discussion

Continuous modulation of Cas12a trans-cleavage activity driven by Hooded DNA

Previous studies have shown that CRISPR/Cas12a exhibits varying trans-cleavage activities toward different DNA probes [23, 25, 26]. However, the properties of the more versatile split CRISPR/Cas12a system in this regard have not yet been reported. Therefore, we designed DNA probes with diverse structures to systematically investigate the trans-cleavage activity of split CRISPR/Cas12a (Fig. 1A and S1).

Fig. 1.

Validation of HAT structure principle and continuous modulation of Cas12a activity. (A) Schematics of split CRISPR/Cas12a substrates with different structures. (B) Trans-cleavage activity of Cas12a toward dsDNA and triplex DNA substrates. Ra: blunt-end probe; Ra-3′: probe with a 3′ overhang; Ra-5′: probe with a 5′ overhang; CPLT: probe with a club-shaped overhang. (C) Trans-cleavage activity of Cas12a toward HAT structures. Single: single-stranded probe; Double: complementary dsDNA probe; HAT: dsDNA probe capped with a HAT structure (30T); HAT-split: HAT structure after being opened. (D) Gel electrophoresis analysis of hairpin, HAT-5T, HAT-15T, and HAT-30T probes before and after CRISPR/Cas cleavage. (E, F) Continuous modulation via the HAT structure. (E) Fluorescence intensity and fitting curve for HAT probes with varying poly-T lengths. (F) Schematic of continuous length tuning of the poly-T segment in the HAT probe

We first investigated the resistance to trans-cleavage of double-stranded DNA (dsDNA) and triplex DNA (T-shaped) substrates. As shown in Fig. 1B, split CRISPR/Cas12a exhibited high trans-cleavage activity, rapidly cleaving dsDNA substrates, while its activity was slightly reduced toward triplex DNA substrates. To further evaluate the effect of terminal structures on trans-cleavage activity, we designed several probe variants based on previously reported configurations: blunt ends (Ra), 3’ overhangs (Ra-3’), 5’ overhangs, and allosteric hairpin-shaped ends. Consistent with earlier findings, split CRISPR/Cas12a preferentially cleaved substrates with 3’ overhangs, while its activity was significantly reduced toward those with 5’ overhangs (Fig. 1B). Further investigation revealed that the Hooded/HAT structure exhibited complete resistance to trans-cleavage, likely due to its bulged bubble-like structure, as disruption of this structure abolished its resistance (Fig. 1C). To explore the influence of HAT structure on reaction kinetics, we designed single-stranded bubble (hairpin) structures and HAT structures of varying sizes. Gel electrophoresis results indicated that single-stranded bubble and “small” HAT structures were completely digested by split CRISPR/Cas12a, whereas “large” HAT structures remained uncleaved, suggesting that the trans-cleavage activity of split CRISPR/Cas12a depends on the size of the HAT structure (Fig. 1D). We therefore engineered a series of Hooded probes with bubble regions comprising poly‑thymine (poly‑T) tracts ranging continuously from 5 to 30 nucleotides (T). Trans‑cleavage efficiency decreased exponentially with increasing poly‑T length, becoming negligible when the T‑count exceeded 25 (Fig. 1E). Based on these results, we demonstrated that by modulating the size of the HAT structure, it is possible to finely and continuously regulate Cas12a activity in a manner analogous to a “molecular rheostat” (Fig. 1F).

Mechanistic basis of Hooded DNA‑mediated inhibition

The inverse correlation between trans-cleavage rate and poly-T length (Fig. 1E) suggested a steric hindrance mechanism. To test this, we performed molecular docking and dynamics simulations to compare the affinity of Cas12a for HAT DNA versus a conventional hairpin substrate. Simulations were based on a Cas12a crystal structure (PDB: 5NFV) with its target DNA removed, which revealed an open conformation with an everted Nuc domain and increased REC2-Nuc distance, facilitating reporter access (Figure S2). Models of HAT and hairpin DNA (each with a 25-nt loop and stem) were generated using AlphaFold3, optimized, and docked into the RuvC catalytic site. Subsequent 100-ns all-atom molecular dynamics simulations showed that the hairpin DNA complex had a significantly lower (more favorable) binding free energy than the HAT DNA complex (ΔG = 38.29 kcal/mol), indicating stronger protein-DNA interactions and higher thermodynamic stability (Fig. 2A). Root-mean-square deviation (RMSD) analysis confirmed stable trajectories for both complexes, with smaller fluctuations observed for the hairpin complex, reflecting higher conformational stability upon binding (Fig. 2B). Interaction analysis further revealed that the hairpin DNA engaged in more numerous and persistent protein-DNA contacts and hydrogen bonds within the RuvC domain compared to HAT DNA (Fig. 2C and S3). Importantly, the HAT architecture contains duplex arms flanking the loop region, which impose additional spatial constraints that limit loop flexibility and reduce its accessibility to the catalytic site. In contrast, the loop of a conventional hairpin is more exposed and conformationally adaptable, rendering it more susceptible to trans-cleavage. Together, these computational results indicate that HAT DNA experiences increased steric occlusion, resulting in reduced engagement with the Cas12a catalytic center.

Fig. 2.

Mechanism of Hooded DNA resistance to Cas12a trans-cleavage. (A) Comparison of binding free energies for Cas12a-HAT and Cas12a-hairpin complexes calculated by MM-GBSA. (B) Backbone RMSD of the two complexes during 100-ns MD simulations. (C) Time-dependent number of protein-DNA contacts for HAT- and Hairpin-bound complexes. (D) Comparison of Cas12a trans-cleavage activity toward hairpin DNA and HAT DNA. (E) Cas12a trans-cleavage activity toward hairpin DNAs with different stem lengths

Fluorescence resonance energy transfer (FRET) assays further corroborated these insights. Activated Cas12a rapidly cleaved hairpin DNA but showed minimal activity against HAT DNA (Fig. 2D). Notably, Cas12a activity toward hairpin substrates was strongly dependent on stem length: hairpins with short duplex stems (10 nt) showed limited cleavage, whereas those with longer duplex stems (25 nt) were cleaved more efficiently (Fig. 2E). Given the known high activity of Cas12a toward dsDNA probes [27], we speculate that short hairpin stems are protected by partial steric effects from the adjacent loop, whereas longer duplex segments are more exposed to the catalytic site and thus more susceptible to cleavage. In contrast, under identical conditions, the HAT architecture provides enhanced steric shielding not only to the loop region but also to the complementary reporter strand. Collectively, these results demonstrate that HAT structures suppress trans-cleavage by increasing steric hindrance, restricting loop accessibility, and simultaneously protecting the complementary probe strand, thereby establishing a robust structural basis for tunable Cas12a activity regulation.

Gating control and optimization of the CRISPR/Cas12a system using Hooded DNA

The robust resistance of Hooded DNA to trans-cleavage makes it an ideal gating element for CRISPR/Cas12a systems, where the presence or absence of the HAT structure dictates an “ON” or “OFF” output. We therefore investigated its gating utility and potential for biosensing applications.

We first examined how structural features of the Hooded DNA affect gating performance. While a 25‑nt HAT region sufficed for near‑complete activity suppression (“OFF” state), efficient activation (“ON” state) depended on the timely displacement of the HAT cap, a process influenced by the length of the non‑complementary spacer region. Systematic variation of the spacer length revealed that spacers of 6 nt or shorter minimally impacted the blocking efficiency but significantly impaired the subsequent cleavage of the opened HAT structure (Fig. 3A and S4). A 5‑nt spacer was identified as the optimal compromise, providing effective blocking while maintaining high post‑activation cleavage activity (Fig. 3B).

Fig. 3.

Gating regulation of CRISPR/Cas12a activity by Hooded DNA. (A) Fluorescence kinetics of HAT-structured probes with varying spacer lengths before and after strand separation. (B) Cleavage activity comparison for different spacer lengths pre- and post-separation. (C) Schematic of the SCHP detection mechanism involving AuNP-enriched probes. (D) TEM images of AuNPs and AuNP-Probe conjugates. (E) Zeta potential of AuNPs and AuNP-Probe. (F, G) Detection performance of the SCHP system. (F) Fluorescence kinetics for AuNP-Probe cleavage at different target miRNA concentrations. (G) Corresponding calibration curve

Moreover, the inherent trans-cleavage activity of the split Cas12a system itself also influences the gating performance and overall biosensing capability. To harness this for sensitive RNA detection, we optimized the split CRISPR/Cas12a system. Evaluation of four clinically relevant miRNAs confirmed reproducible and stable fluorescence responses (Figure S5A). The system demonstrated high sequence specificity, with significant signals only for perfectly matched miRNA‑DNA pairs (Figure S5B). A comprehensive parameter optimization was conducted. Among tested Cas12a orthologs, LbCas12a yielded the highest sensitivity (Figure S5C and D). Optimal conditions were determined as follows: 500 nM Cas12a, 8 mM Mg²⁺, 200 nM scaffold RNA, 200 nM DNA probe, reaction at 39 °C for 60 min in r2.1 buffer (Figure S5E‑K).

Finally, to maximize biosensing capability, we constructed a multiplexed sensing system termed Split‑CRISPR/Cas12a with Hooded DNA Probe (SCHP). In this design, a DNAzyme serves as the HAT‑opening switch due to its programmable cleavage activity. A riboadenosine (rA) site was introduced within the poly‑T region, enabling DNAzyme‑mediated conversion of the HAT to an open conformation. Gold nanoparticles (AuNPs) were employed as carriers for the HAT probe to enhance local concentration and signal amplification (Fig. 3C). Successful probe adsorption onto AuNPs was confirmed by transmission electron microscopy (TEM) and zeta potential measurements (Fig. 3D, E). Using a FAM‑labeled HAT probe, we verified efficient fluorescence quenching by AuNPs and high DNAzyme cleavage activity (Figure S6). The system exhibited a detectable fluorescence response within 150 min at femtomolar miRNA concentrations, with a strong linear correlation between signal and the logarithm of miRNA concentration and a calculated limit of detection (LOD) of 2.05 × 10⁻¹⁵ M (Fig. 3F, G). These findings establish Hooded DNA as an effective gating module for CRISPR/Cas12a, enabling highly sensitive biosensing applications.

Dual-Mode multiplexed detection of protein and MiRNA via an integrated Hooded DNA-CRISPR/Cas12a platform

We next expanded the SCHP system to perform logic‑gated, dual‑mode detection of protein and miRNA biomarkers. As a proof‑of‑concept, we targeted prostate‑specific antigen (PSA) by engineering a DNAzyme‑PSA‑Aptamer (DPA) gate. The DPA construct adopts a hairpin structure that sequesters the DNAzyme catalytic core (Figure S6 and S7). PSA binding triggers conformational unfolding, activating the DNAzyme to cleave an rA‑containing HAT probe and release a fluorescent signal (Fig. 4A). This mechanism was validated by the observation of significant fluorescence only in the presence of all three components: DPA, PSA, and the probe (Fig. 4B).

Fig. 4.

Dual-mode logic detection of PSA and miRNA. (A) Schematic of the DPA reaction mechanism with and without PSA. (B) Validation of probe cleavage by DPA in the presence of PSA. (C) Fluorescence kinetics and calibration for different PSA concentrations. (D) Schematic of the combined CRISPR/Cas12a and DPA detection strategy. (E) Specificity validation of the combined detection. All with scaffold RNA, spacer RNA, Cas12a and HAT-Probe@AuNPs. P.C. (positive control): activated Cas12a with single-stranded reporter

The fluorescence response increased with PSA incubation time, reaching a plateau at approximately 60 min (Fig. 4C). We subsequently optimized key reaction parameters. The fluorescence signal correlated positively with DPA reaction time up to 180 min (Figure S8A). Signal intensity increased gradually with DPA concentrations from 100 to 250 nM and saturated at 300 nM (Figure S8B). Among tested divalent metal ions, Mg²⁺ provided the strongest enhancement compared to Ca²⁺ and Mn²⁺, with an optimal concentration of 8 mM (Figure S8C, D). Fluorescence also increased with probe concentration from 50 to 500 nM, plateauing at 400 nM (Figure S8E). Under these optimized conditions, the system achieved quantitative PSA detection across a concentration range from 10 ng/mL to 1 pg/mL, with a strong linear correlation between fluorescence intensity and the logarithm of PSA concentration (Figure S9).

We then integrated this DPA gate with the split CRISPR/Cas12a system to construct a dual-mode logic sensor for the concurrent detection of PSA and miRNA. In this design, PSA binding opens the DPA hairpin, activating the DNAzyme to cleave and release the HAT probe. Simultaneously, specific target miRNAs bind to their cognate DNA activators, which collectively trigger Cas12a trans-cleavage activity. The activated Cas12a then cleaves the uncapped HAT probe, generating a fluorescence readout (Fig. 4D). This AND-gate logic was confirmed, as a significant signal required the simultaneous presence of both PSA and miRNA (Fig. 4E). Incorporating AuNPs for probe enrichment further improved the sensitivity of this combined assay (Figure S10).

The system demonstrated robust specificity in complex environments. It maintained selective responses to PSA against interfering proteins such as HSA, IgG, and FBS, and functioned effectively in plasma matrices (Figure S11). Specificity for the target miRNAs was also verified against non-target miRNA sequences (Figure S12). Finally, we developed a one-pot assay combining the PSA incubation and miRNA detection steps. Although the reaction kinetics were slower, this simplified format retained clear discrimination from target-free controls (Figure S13), indicating practical diagnostic potential.

Multiplexed SCHP system for precise diagnosis of prostate cancer

PCa is a leading male malignancy with rising global incidence [28]. While serum PSA testing is sensitive for early detection, its specificity is limited due to elevation in benign conditions like prostatitis, leading to false positives [29, 30]. Combining PSA with tumor‑specific biomarkers, such as microRNAs (miRNAs), is a promising strategy to improve diagnostic accuracy. We therefore incorporated a PCa‑specific miRNA panel into our multiplexed SCHP system for logic‑based dual‑mode detection.

By integrating data from the TCGA database and prior literature, we identified five candidate miRNAs upregulated in PCa (Figs. 5A, B). Validation using cell lines and 48 clinical plasma samples led to the selection of miR‑21 and miR‑148 as a specific diagnostic panel, whose plasma average expression levels are more than twice as high in PCa (Figs. 5C, D, S14‑S17).

Fig. 5.

Multiplexed detection for prostate cancer diagnosis. (A) Analysis of highly expressed miRNAs in prostate cancer patients based on TCGA database. (B) Schematic illustration of the screening process for selecting and validating potential highly expressed miRNAs in prostate cancer patients with RT-qPCR from published literature and the TCGA database. (C) PCR validation of miRNA expression in Prostatic epithelial cell line (RWPE-1) and prostate cancer cell lines (C42, PC3, Rv1). (D) PCR analysis of miRNA expression in plasma samples from prostate cancer patients (n = 24:24). (E) Schematic diagram of the process for blood collection from BPH and PCa patients, followed by plasma separation via centrifugation and subsequent extraction of PSA and miRNA. (F) Clinical characteristics of BPH and prostate cancer patients, including group distribution, age, PSA level, and ISUP grade. (G) Comparison of relative fluorescence intensities in the DNAzyme-PSA detection assay. (H) Comparison of PSA quantification between ELISA and the DPA assay. (I) Combined detection of miR-21 and miR-148 using SCHP-based CRISPR/Cas12a assay. (J) ROC curve for the combined detection of miR-21 and miR-148. (K) Scatter plot of combined miRNA and PSA detection in plasma samples from prostate cancer patients. ΔFluorescence intensity were normalized to a maximum value of 1. (L) ROC analysis of the combined miRNA and PSA detection strategy. *, P < 0.05, **, P < 0.01, ***, P < 0.001

To evaluate the diagnostic performance of the multiplex sensing system, we established a clinical liquid biopsy strategy for PCa based on this platform (Fig. 5E) and further assembled a validation cohort comprising plasma samples from 48 patients (Fig. 5F). First, we compared the performance of the DPA-based method with conventional ELISA for PSA detection. The results indicated that the DPA method showed a significant increase in PSA signal in PCa patients (Fig. 5G) and yielded consistent results with ELISA (Fig. 5H). Combined detection of miR-21 and miR-148 via qPCR further confirmed their upregulated expression in the plasma of PCa patients, with the combined panel achieving an AUC of 0.82 (Fig. 5I and J).

Finally, using the SCHP multiplex sensing platform for logic-gated detection of both PSA and miRNA in plasma, we observed overall higher fluorescence signals in PCa patients compared to those with BPH. With a cut-off value set at 0.22, the assay achieved 100% specificity and 79% sensitivity, with an AUC of 0.91 (Fig. 5K and L). These results demonstrate that the developed SCHP platform enables multi-dimensional analysis of plasma samples and supports high-specificity identification of PCa patients.

SCHP system combined with CRISPR/Cas12a for Cell-Specific imaging

To further evaluate the robustness and applicability of the SCHP multiplex sensing platform, we applied it to intracellular imaging. For this purpose, we constructed an imaging system based on a tetrahedral DNA nanostructure (TDN) carrier for the simultaneous detection of uracil DNA glycosylase (UDG) and miR-375 expression levels in cells. As illustrated in Fig. 6A, the HAT probe modified with deoxyuridine (dU) was loaded onto the TDN to form a TDN probe system (TDP). This system, along with the split Cas12a components, was delivered into cells via electroporation. In the presence of highly expressed UDG and APE1 within the cells, the dU sites are sequentially recognized and cleaved, leading to the release of single-stranded probes. When miR-375 is also highly expressed, the split Cas12a system is activated, cleaving the probes and generating a fluorescent signal. This enables multi-target-activated intracellular fluorescence imaging based on the simultaneous presence of both biomarkers.

Fig. 6.

Intracellular activation of CRISPR/Cas12a by DNA tetrahedron-based probes for in situ fluorescence imaging in cells. (A) Schematic illustration of the cellular entry of a tetrahedron-loaded HAT-opening probe via electroporation, followed by intracellular UDG-mediated cleavage and subsequent activation and processing by the CRISPR/Cas12a system. (B) Schematic and validation of the fluorescence generation mechanism from HAT-opening probes cleaved by UDG and APE1 enzymes in vitro. (C) Gel electrophoresis analysis of tetrahedron assembly. (D) Quantitative analysis of cellular fluorescence before and after electroporation. (E) Immunohistochemistry (IHC) analysis of prostate tissues from BPH and PCa patients based on database information. (F) Confocal laser scanning microscopy imaging of DNA tetrahedron probes in normal versus prostate cancer cells

First, the feasibility of the proposed mechanism was verified in vitro. The results showed that significant fluorescence activation was observed only when the HAT probe, UDG, and APE1 were all present (Fig. 6B). Successful assembly of the TDN (Fig. 6C) and the TDP probe system was confirmed by gel electrophoresis (Figure S18). Furthermore, we pre-evaluated the impact of electroporation on cell viability. Fluorescence imaging and quantitative analysis revealed that cell mortality rates in the control group, the electroporation-only group, and the group subjected to electroporation with TDP delivery were all below 5%, with no significant differences among the groups, indicating that the electroporation delivery strategy is highly biocompatible.

Based on immunohistochemistry (Fig. 6E) and miRNA expression data (Figure S19) from the public databases, we selected three cell lines for evaluating the cell type-specific response of the imaging system: normal 293T cells (low expression of both UDG and miR-375), bladder cancer 5637 cells (high UDG, low miR-375), and PCa RV-1 cells (high expression of both UDG and miR-375). Confocal imaging results (Figure S20) demonstrated strong FAM fluorescence signals exclusively in the cytoplasm of RV-1 cells, indicating that the split CRISPR/Cas12a system was specifically activated only in the PCa cells. This outcome confirms the strong cell-type discrimination capacity of the constructed imaging system, providing a foundation for the development of highly specific intracellular imaging strategies for tumor detection.

Conclusion

In summary, this study investigated the trans-cleavage activity of the split CRISPR/Cas12a system toward various substrate probes, leading to the development of a Hooded DNA probe capable of switching and flexibly regulating trans-clease activity. These findings provide valuable insights and extensions for the development and application of split CRISPR/Cas12a systems. Furthermore, leveraging the switching behavior of the Hooded probe, we constructed a multiplex, rapid, and highly specific multimodal detection platform. This system demonstrated excellent specificity in the simultaneous detection of miRNA and PSA, enabling efficient discrimination of PCa patients. Its design is highly versatile and programmable; simple replacement of the DNA spacer or aptamer sequence permits adaptation to diverse disease-related nucleic acids and protein biomarkers. Additionally, we validated its logic-gated imaging capability across multiple cell models, confirming the robustness of the system in complex biological environments. These results underscore the broad application potential of this Hooded probe-based strategy for regulating CRISPR/Cas12a activity. In the future, this platform holds promise for adaptable and efficient detection of diverse disease markers by incorporating alternative recognition units, with broad potential for translation in liquid biopsy, companion diagnostics, and precision medicine. Further integration with detection devices and optimization of signal readout will advance its practical application in point-of-care testing and high-throughput clinical screening.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Xingyu Zhong: Investigation, Methodology, Writing-origin draft. Xi Gong: Investigation, Methodology, Writing-origin draft, Validation. Na Zeng: Methodology, Formal analysis, Data curation. Tianci Xie: Resources, Investigation. Shaogang Wang: Conceptualization, Writing-review & editing. Qidong Xia: Supervision, Writing-review & editing, Funding acquisition. All authors were involved in the preparation of the final manuscript.

Funding

None.

Data availability

The datasets supporting the conclusions of this article are included within the article and its additional file.

Declarations

Ethics approval and consent to participate

The clinical study protocol was approved by the Ethics Committee of Tongji Hospital (TJ-IRB202407023), and informed consent was obtained from all participants.

Competing interests

The authors declare no competing interests.