Modular RCA-CRISPR/Cas12a amplification on a multi-volume SlipChip for ultrafast, single-copy quantification of circRNA and miRNA in ovarian cancer

Abstract

The aberrant expression of RNAs in ovarian cancer (OC) progression highlights their potential as clinical biomarkers. However, rapid and accurate quantification of these RNAs in biosamples remains a significant challenge. In this study, we develop a modular isothermal rolling circle amplification (RCA)-activated Cas12a loop-enhanced (MIRACLE) amplification method for circRNA and miRNA quantification without the need of reverse transcription. In this design, isothermal amplification of modular DNA can be initiated by target-specific RCA primers or miRNAs, with the amplification products subsequently recognized by the Cas12a system to generate measurable signals. When integrated with a multi-volume sliding chip (SlipChip) platform, this MIRACLE method enables portable, rapid and ultra-sensitive quantification of these two types of RNA. Under optimized conditions, this platform exhibits detection limits of 0.125 copies per μL for circRNA and 0.326 copies per μL for miRNA, covering a 5-log dynamic range from 10⁻¹ to 10³ copies per μL within 35 min. The platform was validated using OC cell lines and clinical blood samples. It successfully profiled OC RNA biomarkers (hsa_circ_0049101 and hsa-miR-338-3p) and effectively distinguished between early and advanced stages of OC. These results show a strong correlation with RT-qPCR (R² = 0.953 for circRNA and R² = 0.947 for miRNA). This work establishes a versatile CRISPR-microfluidic platform for cancer diagnosis. Its modular design allows for adaptation to detect other cancer-related RNA biomarkers, thereby addressing critical needs in precision oncology.

---

1. Introduction

According to the Global Cancer Statistics (2023), ovarian cancer (OC) is one of the most lethal gynecological malignancies worldwide.¹ This poor prognosis primarily stems from the fact that over 70% of patients are diagnosed at advanced stages (III or IV).² The absence of specific symptoms during early-stage progression, combined with the limited sensitivity and specificity of current screening methods (e.g., CA125 testing and transvaginal ultrasound), has prevented the establishment of an effective early detection system.^3,4 These challenges highlight the urgent need for novel biomarker detection technologies to boost early diagnosis and patient outcomes. Currently, circular RNAs (circRNAs) and microRNAs (miRNAs), as key members of the non-coding RNA family, have demonstrated significant potential as biomarkers for OC.^5,6 CircRNAs lack the 5′ cap and 3′ poly(A) tail found in linear RNAs, forming a stable closed-loop structure that persists in body fluids. Their ubiquitous presence in human cells and tissue-specific expression patterns further enhance diagnostic utility.^7,8 As short (20–25 nt) non-coding RNAs, miRNAs also excel in early detection; for instance, oncogenic miR-200 family members are actively secreted via exosomes during initial tumorigenesis.^9,10 Notably, these molecules exhibit synergistic regulation: circRNAs acting as “molecular sponges” sequester miRNAs via the ceRNA mechanism, modulating oncogenic processes.^11,12 These advantages not only position circRNA and miRNA as promising biomarkers for early-stage OC but also lay the foundation for developing high-precision liquid biopsy platforms.

Quantitative reverse transcription polymerase chain reaction (qRT-PCR) is the gold standard for RNA detection. However, its application in quantifying circRNAs is limited, as conventional PCR strategies fail to efficiently amplify their distinctive backsplice junctions and are prone to interference from homologous linear RNAs.¹³ Northern blotting lacks sufficient sensitivity to meet the requirements for low-abundance circRNA detection, while high-throughput sequencing faces challenges of high costs and complex data analysis.^14,15 In miRNA detection, qRT-PCR is prone to primer–dimer formation and nonspecific amplification, in addition to the lack of reliable reference genes. Microarray technology has a limited dynamic range and carries the risk of cross-hybridization, whereas miRNA sequencing is affected by sequence bias during library preparation, leading to quantitative inaccuracies.^16–18 These conventional methods generally require large starting sample quantities, involve cumbersome and time-consuming procedures, and fail to enable dynamic monitoring, significantly hindering the application of circRNAs and miRNAs in clinical diagnostics.¹⁹ Therefore, developing highly sensitive and specific RNA detection methods represents an urgent and challenging task. In recent years, nucleic acid detection technologies based on clustered regularly interspaced short palindromic repeats (CRISPR) and CRISPR-associated nucleases (Cas) have advanced rapidly. The CRISPR–Cas system achieves target-specific recognition through programmed CRISPR RNA (crRNA), providing a highly specific platform for nucleic acid detection.²⁰ For example, Cas12a, after forming a complex with crRNA, not only specifically recognizes target DNA but also activates trans-cleavage activity, enabling the cleavage of non-specific single-stranded DNA (ssDNA). This characteristic endows it with remarkable specificity in nucleic acid detection.²¹ Currently, CRISPR/Cas12a has been widely applied in the detection of cancer-related RNAs. For instance, Chen et al.²² developed the DETECTR system, which combines reverse transcription-recombinase polymerase amplification (RT-RPA) with CRISPR/Cas12a to achieve detection and dynamic monitoring of cancer-associated miRNAs. Meanwhile, Ke et al.²³ proposed the RETA-CRISPR system, utilizing RT-RCA coupled with CRISPR/Cas12a for detection of lung cancer-related circRNAs. However, these technologies still have limitations: lack of universality, requiring redesign and optimization of crRNA for each new target, which raises development costs and time. Additionally, the Cas12a system needs reverse transcription for RNA detection which extends processing times and may introduce variability that reduces sensitivity. Future efforts should focus on optimizing these technologies for better universality and convenience.

Herein, we develop modular isothermal RCA-activated Cas12a loop-enhanced (MIRACLE) amplification, a reverse-transcription-free platform integrating a multi-volume SlipChip for ultrasensitive circRNA and miRNA quantification. This system surmounts critical limitations of the gold-standard qRT-PCR and state-of-the-art CRISPR diagnostics (e.g., DETECTR/SHERLOCK) by delivering 10-fold higher sensitivity, completing detection within 30 min, and reducing cost to $1.98 per test. This system overcomes key limitations of conventional approaches through three pioneering innovations: (i) universal circDNA scaffold technology that converts heterogeneous RNA targets into a standardized Cas12a-recognizable template, eliminating redundant assay development; (ii) BstNI nicking enzyme-mediated recognition enabling direct back-splice junction (BSJ) targeting without reverse transcription, circumventing linear RNA interference; (iii) CRISPR-microfluidic cascade amplification achieving single-copy sensitivity via programmable trans-cleavage (signal generation) and cis-cleavage (recognition module recycling). By synergistically merging molecular engineering with tunable microfluidics, MIRACLE establishes a new paradigm for rapid (<30 min), multiplexed liquid biopsy. We anticipate that this platform will accelerate the clinical translation of RNA biomarkers for early cancer diagnostic.

2. Materials and methods

2.1 Reagents and instruments

All HPLC-purified oligonucleotide sequences are listed in Table S1. These oligonucleotides were synthesized by GenScript Biotechnology Co., Ltd. (Nanjing, China). The following reagents were purchased from New England Biolabs (Ipswich, MA, USA): BstNI nuclease, T4 RNA ligase 2, RNase inhibitor, DNase I, RNase R, dNTPs, HiScribe™ T7 High Yield RNA Synthesis Kit, Monarch® RNA Cleanup Kit, and EnGen® LbaCas12a (Cpf1). The GoScript™ Reverse Transcription Kit was obtained from Promega Biotech Co., Ltd. (Beijing, China). CircLigase™ ssDNA ligase was purchased from Lucigen Biotech Co., Ltd. (Madison, WI, USA). Cell lines, including OC cells (NIH:OVCAR-3, SKOV3), normal ovarian cells (IOSE-80), endometrial cancer cells (RL-952), breast cancer cells (MCF-7), cervical cancer cells (HeLa), and human embryonic kidney cells (HEK293T), were purchased from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). Human carbohydrate antigen 125 (CA125) ELISA Kit, and human epididymis protein 4 (HE4) ELISA Kit were sourced from Sangon Biotech Co., Ltd. (Shanghai, China).

2.2 Design and synthesis of circDNA scaffold

The modular circDNA scaffold was engineered with interchangeable recognition modules targeting specific circRNA and miRNA sequences, enabling plug-and-play adaptability through sequence replacement. Synthesis was performed in a 20 μL reaction system containing 50 pmol HPLC-purified ssDNA template (GenScript), 5 U CircLigase™ ssDNA Ligase (Lucigen, Madison, WI, USA), and 8 μL 2.5× CircLigase Buffer 2, followed by 60 °C incubation (1 h) and 85 °C heat inactivation (10 min). Residual linear templates were eliminated via 40 U Exonuclease I (Thermo Fisher, Waltham, MA, USA) treatment (37 °C, 15 min), with subsequent purification using Monarch® PCR Cleanup Kit. Final circDNA products were quantified by a Qubit™ 4.0 Fluorometer and stored at −80 °C until use.

2.3 The MIRACLE assay

All frozen RNA samples were thawed on ice prior to use, with aliquots immediately returned to −80 °C storage after removal. Biological reagents were thawed on ice while chemical solutions were equilibrated to room temperature. The experimental workflow for circRNA detection consisted of BstNI nicking reaction (10 μL system): 1 μL serially diluted circRNA template, 1 μL RNase inhibitor, 5 μL BstNI (6 U μL⁻¹), 2 μL pre-RCA primer (100 nM), 1 μL MgSO₄ (100 mM). RCA amplification: 2 μL circDNA scaffold (40 nM), 1 μL dNTP mix (2.5 mM each), 1 μL φ29 DNA polymerase (2 U μL⁻¹), 2 μL 10× reaction buffer, nuclease-free water to 20 μL final volume, followed by incubation at 48 °C for 15 min. CRISPR/Cas12a detection: the RCA products were mixed with 2 μL crRNA (200 nM), 2 μL LbCas12a (200 nM), 1 μL ssDNA reporter (400 nM), and 1 μL DTT (10 mM) and incubated at 37 °C for 30 min. Endpoint fluorescence was measured by an Infinite® 200 PRO microplate reader, fluorescence kinetics was monitored by a QuantStudio™ 3 Real-Time PCR system. For miRNA analysis, step BstNI was omitted, while all other procedures remained consistent with the aforementioned protocol.

2.4 Blood sample collection

Clinical plasma samples, including 24 healthy controls and 24 ovarian cancer cases, were collected from Liaoning Cancer Hospital (Shenyang, China). All participants (aged 20–75 years) provided written informed consent. Ovarian cancer patients were pathologically confirmed and had not received preoperative anticancer therapy. The detailed clinical information of OC patients and healthy volunteers, including pathological types, TNM stages, and age, is provided in Table S2. For sample processing, 5 mL of peripheral blood was collected intravenously into EDTA anticoagulant tubes, gently inverted for mixing, flash-frozen on dry ice, and stored at −80 °C for subsequent experiments. The study protocol (Approval No. DUTSBE240612-02) was approved by the Ethics Committee of Dalian University of Technology and adhered to the Declaration of Helsinki guidelines.

2.5 RNA extraction from cell lines and whole blood

Total RNA extraction from cell lines and whole blood samples was performed using the miRNeasy Micro Kit (Qiagen #217084) and TRIzol® Plus RNA Purification Kit (Thermo Fisher Scientific #12183555) for simultaneous isolation of circRNA and miRNA. The standardized workflow comprised the following steps.

RNA extraction from cell lines. Ovarian cancer cell lines (NIH:OVCAR3, SKOV3) and normal ovarian epithelial cells (IOSE-80) were cultured as detailed in the SI: cell lines and culture conditions. Upon reaching 80% confluence, cells were washed twice with 1× PBS and lysed with 1 mL TRIzol reagent per well (6-well plate, seeding density: 1 × 10⁶ cells per well) for 5 min at room temperature. Lysates were transferred to RNase-free tubes. For circRNA enrichment, total RNA (5 μg) was incubated with 3 U μg⁻¹ RNase R at 37 °C for 15 min. Enzyme inactivation: 70 °C for 15 min, circRNA recovery: centrifugation at 12 000g for 15 min. For miRNA isolation: parallel samples were processed using an miRNeasy Micro Kit following the manufacturer's protocol.

For blood RNA extraction. Human plasma samples obtained from the hospital were thawed on ice following standard venipuncture procedures. Total RNA isolation was performed using the miRNeasy Micro Kit according to the manufacturer's protocol. circRNA and miRNA were isolated separately from 500 μL plasma. For circRNA, total RNA extracted via miRNeasy Micro Kit underwent RNase R digestion (3 U μg⁻¹, 37 °C for 15 min), heat-inactivation, and centrifugation before elution in 14 μL of RNase-free water. For miRNA, samples spiked with cel-miR-39 carrier RNA were processed identically (excluding RNase R step) and eluted in 30 μL of RNase-free water DNase treatment was performed on-column for all samples (Qiagen #79254).

RNA concentration and integrity were measured using a NanoPhotometer® N60 ultra-micro spectrophotometer (Implen GmbH, Germany). Qualified samples were stored at −80 °C ultra-low temperature conditions.

2.6 RT-qPCR detection RNA

Reverse transcription was performed using the GoScript™ Reverse Transcription System (Promega, USA) with target-specific primers to convert total RNA (1 μg pretreated with RNase R) into cDNA. For miRNA quantification, cDNA synthesis utilized stem-loop primers (10 μM) and M-MLV Reverse Transcriptase (Promega) at 42 °C for 60 min, followed by enzyme inactivation at 70 °C for 15 min.

Quantitative PCR reactions were performed in a final volume of 20 μL containing 2 μL cDNA template, 0.4 μL forward primer (10 μM), 0.4 μL reverse primer (10 μM), and 10 μL GoTaq® qPCR Master Mix (Promega), with nuclease-free water added to adjust volume. Amplification conditions comprised an initial denaturation at 95 °C for 10 min, followed by 40 cycles of denaturation at 95 °C for 15 s and combined annealing/extension at 60 °C for 60 s using SYBR Green chemistry. GAPDH and U6 snRNA served as endogenous controls for circRNA and miRNA, respectively. Relative expression was calculated via the ΔCt method. All primer sequences are listed in Table S1.

2.7 MIRACLE SlipChip detection RNA

The MIRACLE SlipChip detection was performed using thawed RNA samples that were maintained on ice during handling and promptly returned to −80 °C after use. The upper and lower chips were assembled in an oil phase consisting of tetradecane and dimethyl silicone oil (1 : 1 v/v). The RNA sample was combined with the BstNI nicking mix (10 μL: 1 μL RNA, 1 μL RNase inhibitor, 5 μL BstNI [6 U μL⁻¹], 2 μL pre-RCA primer [100 nM], 1 μL MgSO₄ [100 mM] in 1× NEBuffer 3.1) at 48 °C for 5 min and 95 °C for 30 s, and then RCA mix (10 μL: 2 μL circDNA scaffold [40 nM], 1 μL dNTPs [2.5 mM each], 1 μL phi29 DNA polymerase [2 U μL⁻¹], 2 μL 10× reaction buffer) was added. The reaction reagents were injected through continuous “pearl-chain” microfluidic channels to form microdroplets in the upper chip wells. Droplets formed in upper chambers were transferred via sliding alignment to lower chambers for 15 min incubation at 48 °C. CRISPR/Cas12a detection mix (2 μL crRNA [200 nM], 2 μL LbCas12a [200 nM], 1 μL ssDNA reporter [400 nM], 1 μL DTT [10 mM]) was then injected, with secondary sliding enabling droplet merging for 15 min reaction at 37 °C. Fluorescence imaging was performed before and after incubation using a Nikon Ti2 microscope, with automated image stitching and analysis via NIS-Elements software (v5.01). The copy numbers were calculated based on most probable number (MPN) likelihood function theory.

2.8 SlipChip design and fabrication

Detailed descriptions of the fabrication and operational protocols for the SlipChip microfluidic device are provided in the SI, including specifications of the materials used, fabrication process, assembly procedures, and detection workflow.

2.9 Data analysis

The statistical analyses in this study were performed using IBM SPSS Statistics 26.0 (IBM, USA) and GraphPad Prism 9.0 (GraphPad Software, USA), with data visualization implemented through GraphPad Prism and Origin 2022 (OriginLab, USA). All quantitative results were derived from ≥3 independent experimental replicates and expressed as mean ± standard deviation (SD). For significance analysis and intergroup comparisons, differences between groups were analyzed using two-tailed Student's t-test or Welch's corrected t-test, with statistical significance defined as p < 0.05. For multiple comparisons, one-way analysis of variance (ANOVA) with Bonferroni correction was applied. Results were annotated with asterisks or explicit p-values in figures. Receiver operating characteristic (ROC) curves were generated to evaluate the diagnostic ability of hsa_circ_0049101 and hsa-miR-338-3p biomarkers in tumor detection. The area under the curve (AUC) was used to quantify predictive power (interpretation: AUC >0.5 indicates statistical significance; AUC = 1.0 represents perfect prediction). The optimal diagnostic cutoff value was determined using the maximum Youden index, with simultaneous calculation of sensitivity, specificity, and other key parameters.

3. Results and discussion

3.1 Design and working principle of the MIRACLE SlipChip platform

To achieve faster, simpler, and more cost-effective detection of target RNAs, we developed MIRACLE (modular isothermal RCA-activated Cas12a loop-enhanced amplification), a reverse transcription-free method for rapid and low-cost detection of circRNAs and miRNAs. It integrates with a multi-volume SlipChip microfluidic platform for quantification (Fig. 1).

Fig. 1 The schematic representation of MIRACLE SlipChip platform for circRNA and miRNA detection. (A) Module DNA design, (B) the principle of MIRACLE assay, (C) the workflow of the MIRACLE SlipChip detection platform.

Specifically, for the MIRACLE method, we designed an innovative pre-RCA primer with three key features (Fig. 1A and B): (a and d) terminal blocking bases to prevent non-specific amplification, (b) a BSJ-binding sequence for specific circRNA recognition, and (c) an embedded BstNI recognition site. Upon binding to the circRNA BSJ, BstNI cleaves this site to release a short mature RCA primer by heat denaturation, initiating downstream RCA. We further engineered a circDNA scaffold containing two functional domains: (1) a target-specific recognition module for primer binding and (2) a universal detection module compatible with CRISPR/Cas12a. This scaffold serves as a template for phi29 DNA polymerase-driven RCA primed by the mature RCA primer or miRNA. The Cas12a RNP complex recognizes the detection module within RCA products, leveraging dual cleavage activities for synergistic amplification: cis-cleavage processes RCA amplicons into secondary primers that trigger new RCA cycles (cascade amplification); trans-cleavage degrades ssDNA reporters, generating quantifiable fluorescence for real-time/endpoint detection.

We further integrated the MIRACLE method with a multi-volume SlipChip microfluidic platform for quantitative RNA biomarker detection in OC cells and blood samples (Fig. 1C). After nucleic acid extraction, samples were incubated with BstNI reaction mix to perform site-specific nicking. The nicked products were then heat-denatured (95 °C, 30 S) and subsequently mixed with RCA mix prior to loading onto the SlipChip microfluidic device for RCA amplification. A pre-assembled CRISPR/Cas12a premix was then injected into the chip. Through a single sliding operation, deterministic droplet merging occurred between the Cas12a reagent array and the amplified product array. Following brief isothermal incubation (<30 min), target-specific fluorescence signals emerged in microwells. This design leverages slip-induced microfluidic alignment to eliminate manual amplicon transfer, significantly reducing cross-contamination risks while accelerating total detection time to <30 min.

3.2 Feasibility analysis of key reactions in the MIRACLE method

OC-related circRNA biomarkers (hsa_circ_0049101, circMUC16) were selected as targets to develop the MIRACLE method and investigate its underlying mechanisms. Inspired by Murray et al., the restriction enzyme BstNI is employed as a nicking enzyme to cleave DNA strands within RNA:DNA heteroduplexes.^24,25 We analyzed BstNI's activity across circRNA and pre-RCA primer complexes, with untreated counterparts serving as controls (Fig. 2A). Electrophoretic analysis showed a 15 nt DNA fragment (RCA primer) exclusively in BstNI-treated samples (lane 5), with no degradation products observed in untreated controls (lane 6). This result demonstrates BstNI selectively nick DNA within circRNA and pre-RCA primer at recognition sites. Then, the template functionality of circDNA scaffolds was validated through comparative RCA experiments employing two primers: (1) a synthetic RCA primer (lane 1) and (2) the RCA primer generated from BstNI nicking (lane 2) (Fig. 2B). Both primers can initiate RCA reaction (lanes 3–6), confirming that BstNI-generated RCA primers retain functionality for subsequent amplification.

Fig. 2 Feasibility analysis of key reactions in the MIRACLE method. (A) The feasibility of BstNI digestion in cleaving circRNA:DNA was verified by 20% PAGE. Lane m dl2000 marker, lane 1 nicked RCA primers purified by gel extraction, lane 2 circRNA, lane 3 pre-RCA primer, lane 4 synthesis RCA primer, lane 5 BstNI + circRNA : DNA (200 nM circRNA : 100 nM pre-RCA DNA primer), lane 6 untreated circRNA : DNA (200 nM circRNA : 100 nM pre-RCA DNA primer). (B) 2% agarose electrophoresis image for the RCA process. Lane m dl2000 marker, lane 1 synthesis RCA primer, lane 2 RCA primer generated by BstNI nicking, lanes 3 and 4 RCA product by synthesis RCA primer, lanes 5 and 6 RCA primer by RCA primer generated by BstNI nicking. (C) Real-time detection of serial 10-fold dilutions of circRNA from 10 fM to 1 nM with 200 nm Cas12a RNP. Control assays were conducted with RNP (Cas12a/crRNA) left out, NC was without circRNA. Error bands represent mean ± s.d. for 3 technical replicates. (D) Gel analysis of the products from the reactions in (C). Lane m dl500 marker, lane 1 10 fM, lane 2 100 fM, lane 3 1 pM, lane 4 10 pM, lane 5 1 nM, lane 6 100 pM with Cas12a rnp, lane 7 1 nM, lane 8 100 pM control without Cas12a RNP. (E) The end-point fluorescence intensity of the MIRACLE assay by different reaction components. Error bands represent mean ± s.d. for 3 technical replicates.

We systematically analyzed MIRACLE activity across circRNA concentrations (10 fM–1 nM). The real-time fluorescence showed a dependent increase with circRNA concentrations ranging from 10 fM to 10 pM, but exhibited a sharp decrease at 100 pM and 1 nM (Fig. 2C). Corresponding gel electrophoresis analysis of the RNP (Cas12a/crRNA) reaction system (Fig. 2D) demonstrated minimal detection of long-chain RCA products at low RNA concentrations (10 fM–10 pM), with faint low-molecular-weight DNA bands observed in lanes 1–4. This observation suggests that Cas12a-mediated cis-cleavage remains functionally active even under limited RCA amplicon abundance. In contrast, long-chain RCA products were observed at 100 pM–1 nM (lanes 5 and 6), with weaker band intensity and notable smearing compared to the control group (lanes 7 and 8). The smearing pattern indicates reduced efficiency of Cas12a-mediated cis-cleavage under high RCA amplicon loads, with trans-cleavage becoming predominant. Cas12a's nonspecific trans-cleavage triggers extensive off-target cutting, depleting substrates and compromising enzymatic activity to suppress cis-cleavage. Consequently, no cleaved amplicon fragments were detected.

Meanwhile, high target concentrations trigger rapid RCA reactions, producing ssDNA amplicons that vastly outnumber the reporter probe. This drives Cas12a to preferentially cleave high-copy RCA amplicons. In Fig. S1, in vitro competition assays reveal that at elevated amplicon-to-reporter ratios, Cas12a shifts its cleavage priority from reporters to RCA amplicons, effectively suppressing fluorescence signals. Maintaining circDNA scaffolds at 1/10 reporter concentration minimizes Cas12a-mediated cleavage interference, optimizing RCA–reporter equilibrium to direct trans-cleavage toward fluorescent probes for precise RNA quantification.

Systematic component omission analysis revealed that negative controls (lacking circRNA) and the absent components group exhibited baseline fluorescence levels (Fig. 2E). In contrast, the complete reaction system exhibited exponential fluorescence enhancement. Notably, the lack of any single key component completely inactivated the cascade amplification pathway, thereby confirming the indispensable role of all functional elements in the MIRACLE system.

3.3 Comparison of the kinetics and detection sensitivity of RCA, CRISPR, one-step and multi-step MIRACLE assays

In order to assess the lower detection capabilities of various methods for target analytes and select the optimal solution for specific applications, we further conducted a comparative analysis of the sensitivity of the RCA-only, CRISPR-only, one-pot, three-step, two-step assay for circRNA, and two-step assay for miRNA (Fig. 3A).

Fig. 3 Comparison of the kinetics and detection sensitivity of RCA, CRISPR, one-step and multi-step MIRACLE assays. (A) Comparison of the kinetics and detection sensitivity of the RCA, CRISPR, one-step MIRACLE, and two step MIRACLE assays detection circRNA and miRNA. (B) Real-time fluorescence intensity of the RCA assay. (C) Real-time fluorescence intensity of the CRISPR assay. (D) Real-time fluorescence intensity of the one-step MIRACLE assay. (E) Real-time fluorescence intensity of the three-step MIRACLE assay. (F) Real-time fluorescence intensity of the two-step MIRACLE assay for circRNA. (G) Real-time fluorescence intensity of the two-step MIRACLE assay for miRNA. Error bands represent mean ± s.d. for 3 technical replicates.

RCA-only and CRISPR-only assay detection showed a LOD of ∼100 pM for circRNA (Fig. 3B and C). However, the one-pot assay achieved a sensitivity of 10 fM for circRNA but exhibited lower fluorescence intensity and sensitivity compared to the three-step method, likely due to suboptimal enzyme coordination and system inefficiency (Fig. 3D). The three-step assay, which sequentially combines BstNI-mediated recognition, RCA amplification, and CRISPR/Cas12a, achieved a remarkable LOD of 100 fM, aligning with the reported three-step system (Fig. 3E).²⁶

Based on this, we developed a simplified two-step method, integrating BstNI recognition and RCA into a pre-amplification step, followed by CRISPR/Cas12a fluorescence readout. This streamlined workflow maintained sensitivity and speed comparable to the three-step method (Fig. 3F and G). Further evaluation of the two-step MIRACLE assay demonstrated a detection sensitivity achieving a LOD of 0.1 fM, with no statistically significant difference compared to controls (Fig. S2).

We systematically evaluated the overall amplification efficiency by calculating the number of fluorescent probes generated per unit RNA template (i.e., amplification fold). In a 30 min reaction, the two-step method achieved 10⁴-fold amplification for 10 fM and 100 fM circRNA (Fig. S3). These values exceed the efficiency of phi29 polymerase-driven linear RCA amplification (reported as ∼1000–2000 copies per hour per circular template), primarily due to Cas12a-mediated signal cascade amplification.^27–29 In conclusion, comparative studies of detection modalities demonstrate that synergistic coupling of Cas12a's cis- and trans-activities in our system drives exponential RCA amplification and fluorescence signal amplification simultaneously. These results validate the successful development of a rapid, ultrasensitive method for circRNA and miRNA detection.

3.4 Optimization and specificity of the MIRACLE assay

As discussed above, we successfully established a two-step assay for circRNA and miRNA detection and subsequently optimized key components of the system. The reaction system was optimized through systematic adjustment of critical parameters, including BstNI concentration (Fig. 4A) and reaction temperature (Fig. 4B), pre-RCA primer concentration (Fig. 4C), circDNA scaffold levels (Fig. 4D), phi29 DNA polymerase dosage (Fig. 4E), LbCas12a enzyme quantity (Fig. 4F), crRNA-to-target ratio (Fig. 4G), and ssDNA fluorescent probe concentration (Fig. 4H). This comprehensive optimization led to significantly enhanced catalytic efficiency and accelerated overall reaction kinetics. The optimized MIRACLE detection system achieved peak performance at defined parameters: BstNI (6 U μL⁻¹), pre-RCA primer (100 nM), nicking temperature (48 °C), circDNA scaffold (40 nM), phi29 DNA polymerase (2 U μL⁻¹), LbCas12a/crRNA complex (200 nM each), and ssDNA probe (400 nM). This precision-engineered platform demonstrated ultrasensitive RNA detection capability down to fM levels with single-nucleotide discrimination accuracy.

Fig. 4 Optimization and specificity of miracle assay. (A) BstNI RNAse concentration: 0.05, 0.1, 0.2, and 0.4 U μL⁻¹. (B) BstNI reaction temperature: 28, 38, 48, and 58 °C. (C) Pre-RCA primer concentration: 0.5, 5, 50, and 500 nM. (D) circDNA concentration: 0.5, 5, 50, and 500 nM. (E) phi29 DNA polymerase concentration: 0.5, 1, 2, and 4 U μL⁻¹. (F) Cas12a concentration: 100, 200, 400, and 800 nM. (G) crRNA concentration: 10, 25, 50, and 100 nM. (H) ssDNA probe concentration: 0.02, 0.1, 0.5, and 2.5 nM. (I) Mismatch sequence and corresponding bar diagram by one-base to twelve-base mismatched pre-RCA primer. (J) Fluorescence intensity histogram for the evaluation of anti-interference ability by a mixture of circRNA and linear RNA. (K) Fluorescence intensity histogram for different concentrations of circRNA in different cell matrixes. The mean value (n = 3) with the corresponding standard deviation was reported for each measurement.

Under optimized conditions, we systematically validated the specificity, interference resistance, and universality of this method. For specificity evaluation, the pre-RCA primer-based detection could discriminate single-base mismatches, with the fluorescence intensity of perfectly matched targets being 2-fold higher than M-1 and 6-fold higher than M-8 targets (Fig. 4I). Regarding interference resistance, the circRNA detection efficiency remained stable even when mixed with linear RNA at a 1 : 500 ratio, showing no significant differences between mixed groups (Fig. 4J). Cell-specific analysis revealed high circMUC16 expression in NIH:OVCAR3 cells but low expression in SKOV3 cells (Fig. 4K). NTC reactions, containing all reagents except template DNA, confirmed the absence of contamination. For miRNA detection, MIRACLE exhibited robust specificity: signals from single-base mismatches and four non-target miRNAs (1 pM) were consistently below 2% of the target signal (Fig. S4). Notably, miR-338 was significantly downregulated in ovarian cancer cell lines (NIH:OVCAR3, SKOV3) compared to non-malignant IOSE-80 controls (p < 0.01). Comprehensive validation demonstrates that the MIRACLE system achieves high-fidelity recognition and interference resistance in complex matrices.

3.5 Design and fabrication of the SlipChip

To address the challenge posed by the significant concentration range difference of circRNA and miRNA in human body fluids (high expression: 10³–10⁵ copies per μL; low expression: 10¹–10² copies per μL, spanning three orders of magnitude),^30–32 this study developed a multi-chamber sliding microfluidic chip (SlipChip) with gradient volume self-adaptive capability. The chip system employs a modularly designed 4-level microchamber array (1–125 nL) to achieve dynamic sample volume adaptation. This multi-volume chip consists of upper and lower layers, each containing four distinct volume specifications (1 nL, 5 nL, 25 nL, and 125 nL). Among them, the larger volumes (125 nL and 25 nL) capture low-abundance targets, while the smaller volumes (5 nL and 1 nL) analyze high-concentration samples, with each volume corresponding to 64 chambers (Fig. 5A). The fabricated SlipChip is characterized by comprising two 90 mm × 90 mm upper and lower layers, with specific chip dimensions detailed in the SI (Fig. S5).

Fig. 5 Evaluation of the analytical performance of MIRACLE SlipChip assay. (A) 3D schematic diagram of the MIRACLE Sliphip architecture. (B) Operational workflow demonstrating (i) orange dye injection for microchannel visualization; (ii) first-stage droplet transfer; (iii) blue dye loading; (iv) controlled droplet fusion (yellow/blue = 1 : 1 v/v), with insets showing bright-field micrographs of interfacial dynamics. (C) End-point fluorescence images of MIRACLE SlipChip assay performed on a rotational SlipChip for synthetic hsa_circ_0049101 at five different concentrations. Control, containing no RNA template. (D) End-point fluorescence images of MIRACLE SlipChip assay performed on a rotational SlipChip for synthetic hsa-miR-338-3p at five different concentrations. Serial dilution of RNA template from 0.1 to 1.0 × 10³ copies per μL. Control, containing no RNA template. (E) For each dilution, the approximate contributions of the results from each well volume toward calculating the final concentration were calculated based on the contributions of each volume to the standard deviation, σ (eqn (2)). (F) The linear relationship between the input concentration and the calculated concentration (log Y = 0.804 log X + 0.694 (r² = 0.995)) of hsa_circ_0049101. (G) The linear relationship between the input concentration and the calculated concentration (log Y = 0.882 log X + 0.475 (r² = 0.997)) of hsa-miR-338-3p. Error bands represent mean ± s.d. for 3 technical replicates.

Integrated with the aforementioned MIRACLE method, the chip initially demonstrates the detection workflow through droplet transfer and fusion of differently colored dyes at different steps (Fig. 5B). A clockwise rotation of the top substrate by approximately 10° transfers the liquid to the lower chamber, initiating RCA-mediated target amplification. Upon completion of amplification, the pre-prepared Cas12a cleavage system is injected into the upper chamber. The sliding mechanism enables precise droplet fusion through inter-chamber alignment, achieving volume-based fusion between RCA amplicon droplets and CRISPR/Cas12a premix droplets. Under isothermal conditions, the CRISPR/Cas12a system specifically recognizes target amplicons via crRNA guidance, cleaving ssDNA reporters to release fluorescent signals. Positive samples exhibit distinct fluorescence in their respective chambers. This enclosed operational sequence maintains continuous chamber sealing throughout the process, effectively preventing amplicon exposure to airborne contaminants.

The optimized surfactant coating (Triton X-100/Tween 20) synergistically combined with BSA effectively stabilized the reaction system, overcoming microfluidic confinement effects and non-specific adsorption issues (Fig. S6). EP tube validation confirmed that 0.05% Triton X-100, 0.1% Tween 20, and 1 mg mL⁻¹ BSA preserved detection sensitivity (p > 0.05), enabling robust chip integration for downstream applications.

3.6 Sensitivity and specificity of the MIRACLE SlipChip assay

Using the MIRACLE SlipChip assay, we detected synthetic circRNA and miRNA templates across five log₁₀ dilutions (from 10⁻¹ to 10⁵ copies per μL).

The calculation method for the MIRACLE SlipChip in this study was adapted from previous literature.^33,34 Specifically, RNA concentration determination in this method was based on the most probable number (MPN) likelihood function theory. The concentration λ (copies per μL) was determined by substituting experimental results from each tested volume (i = 1, 2, 3, …) into eqn (1), where n_i, k_i, and v_i represent the total number of chambers, the number of positive chambers, and the chamber volume (μL), respectively, for each volume i. The relative contributions of each chamber's volume are determined by calculating the total uncertainty σ using eqn (2), which was established based on Fisher information theory:

Fig. 5C and D display the endpoint fluorescence images of circRNA and miRNA detection using the MIRACLE SlipChip method. The proportion of positive chambers in each volume group was statistically analyzed, and the template concentration was calculated using formula (1). In negative control experiments, no false-positive results were observed after amplification. As the RNA template concentration increased, the number of fluorescence-positive chambers correspondingly rose. The measured concentration gradients for circRNA were 1.296 × 10³, 1.124 × 10², 1.98 × 10¹, 4.02 × 10⁰, and 0.455 copies per μL (Fig. 5C), while those for miRNA were 1.356 × 10³, 2.357 × 10², 1.23 × 10¹, 1.23 × 10⁰, and 0.128 copies per μL (Fig. 5D). The limits of detection (LODs) were calculated using the formula LOD = 3.3σ/S, where σ represents the standard deviation of 10 no-template controls (NTCs) and S denotes the slope of the calibration curve. This yielded LOD values of 0.125 copies per μL for circRNA and 0.326 copies per μL for miRNA. Strong linear correlations were observed between input and calculated concentrations of hsa_circ_0049101 (log Y = 0.804 log X + 0.694 (R² = 0.995)) (Fig. 5F) and hsa-miR-338-3p (log Y = 0.882 log X + 0.475 (R² = 0.997)) (Fig. 5G). Notably, experimental data across all gradients closely matched the ddPCR-predicted theoretical distribution (Fig. S7).

The MIRACLE SlipChip demonstrates significant advantages over the gold-standard RT-qPCR and state-of-the-art CRISPR platforms (Cas12a-based DETECTR and Cas13a-based SHERLOCK) reported in recent literature. It achieves >10-fold higher sensitivity than qPCR and most CRISPR/Cas methods at the detection limit, while completing the entire assay within 30 min—reducing detection time by 2.4–4.8-fold (Table 1). Cost–benefit analysis (Table 2) reveals a per-test cost of $1.98. Despite utilizing more enzymes, this represents 61.6% savings in total reagent costs versus RT-qPCR ($5.15) and 12–50.5% reduction versus SHERLOCK and ligation-mediated PCR ($2.25–4.00). Critically, the universal crRNA design limits the cost increment to $0.04 per additional target. These advantages originate from: (i) a reusable SlipChip, (ii) avoidance of repeated synthesis via universal crRNA architecture, collectively slashing operational costs.^35–42

Table 1 Performance comparison between MIRACLE Slipchip and recent CRISPR-based detection methods

Method	Application	LOD (copies per μL)	Time (min)	Specificity (%)	Ref.
MIRACLE SlipChip	circRNA/miRNA	0.125/0.326	35	100	This work
DETECTR (Cas12a)	HPV16/18	7.5	60	95	21
SHERLOCK (Cas13a)	Zika virus	10–30	90	93	35
MIRA-CRISPR/Cas12b	Mpox virus	4	45	97	36
RT-RPA-CRISPR	SARS-CoV-2	0.1	40	96	37
PddCRISPR/Cas13a	circRNA	75.3	60	—	38
CRISPR/Cas12a HCR	miRNA	1.204	40	—	39
CRISPR/Cas12a ECL	miRNA	0.765	60	—	40
CRISPR/Cas13a	circRNA	25 × 10^3	60	99	41
qPCR (gold standard)	Broad	1–5	120	99	42

---

Table 2 Cost comparison of MIRACLE Slipchip, RT-qPCR, ligation-mediated PCR, and SHERLOCK v2 methods cycle cost per test (USD) for RNA detection methods

Cost component	MIRACLE SlipChip	RT-qPCR	Ligation- mediated PCR	SHERLOCK v2
Reagents and consumables	$0.85	$1.45	$1.80	$1.25
- Core enzymes	$0.75	$0.80	$1.10	$0.75
- Probes/primers	$0.13	$0.65	$0.70	$0.50
Equipment	$0.60 (fluorescence microscope)	$2.5 (qPCR instrument)	$0.70 (thermocycler)	$0.25 (heat block)
Labor and time	$0.30 (30 min)	$0.90 (120 min)	$1.10 (150 min)	$0.50 (90 min)
QC/Reagent validation	$0.20	$0.30	$0.40	$0.25
Total cost per test	$1.98	$5.15	$4.00	$2.25

---

Within the dynamic range of this MIRACLE SlipChip assay, chambers with distinct volumes exhibit differential contributions to the calculated RNA concentration (Fig. 5E). As the concentration of the control RNA template increases, the dominant contribution to the final calculated concentration shifts from large-volume chambers (125 nL) to medium-volume chambers (25 nL and 5 nL), and then to small-volume chambers (1 nL). The relative contribution of each chamber volume was quantified by its percentage weighting in the total uncertainty (σ) derived from eqn (2). The above results demonstrate that this MIRACLE SlipChip method exhibits remarkable capabilities in detecting low RNA concentrations. Through its multi-volume design, the system significantly expands the dynamic range for RNA detection while maintaining high specificity and excellent interference resistance.

3.7 Analytical performance of circRNA and miRNA in cell and clinical samples

To evaluate the clinical application potential of the MIRACLE SlipChip assay, this study conducted parallel comparative experiments between the MIRACLE SlipChip assay and RT-qPCR using OC-associated culture cell and different-stage clinical blood samples (Fig. 6A).

Fig. 6 MIRACLE SlipChip analysis in cultured cells and clinical sample. (A) Workflow of RT-qPCR assay and MIRACLE SlipChip for testing of the cultured cell and clinical samples. (B) Heatmap of the expression levels of circRNA and miRNA in the cell with OC (NIH:OVCAR3, SKOV3), and control cell (IOSE-80) measured by MIRACLE SlipChip assay. (C) Heatmap of the expression levels of circRNA and miRNA in the blood from patients with OC (n = 24) and healthy donors (n = 24) measured by MIRACLE SlipChip assay. Each sample was tested in two technical replicates. (D) Comparison of the hsa_circ_0049101 levels of NIH:OVCAR3, SKOV3 and IOSE-80 cells determined by MIRACLE SlipChip and RT-qPCR analyses. Comparison of the (E) hsa_circ_0049101, and (F) hsa-miR-338-3p levels of stage iv, iii, stage early and healthy sample determined by MIRACLE SlipChip and RT-qPCR analyses. Error bars, s.d. (n = 3). ****p < 0.0001, ***p < 0.001, **p < 0.01, *p < 0.05, NS (not significant). Ct = 40 indicates undetected (limit of detection: 40 cycles, ND). Note: cross-method correlation shown for trend alignment (non-numerical equivalence). (G) Correlation between the parallel measurements by MIRACLE SlipChip and RT-qPCR for hsa_circ_0049101. (H) Correlation between the parallel measurements by MIRACLE SlipChip and RT-qPCR for hsa-miR-338-3p. (I) ROC curves and AUC analysis of individual hsa_circ_0049101 and hsa-miR-338-3p for the whole stage ovarian cancer diagnosis. (J) ROC curves and AUC analysis of hsa_circ_0049101 and hsa-miR-338-3p for early-stage ovarian cancer diagnosis.

Firstly, we performed parallel analyses of circRNA and miRNA expression profiles using the MIRACLE SlipChip and RT-qPCR in OC cell lines (NIH:OVCAR3, SKOV3) and the control epithelial cell line IOSE-80 (Fig. 6B). hsa_circ_0049101 and hsa-miR-338-3p expression levels in NIH:OVCAR3 and SKOV3 cells were significantly higher than those in control cells, with lower RNA expression observed in SKOV3 cells, consistent with literature reports.^43,44 These biomarkers were all undetectable in the control cells. Notably, the proposed method exhibited enhanced detection sensitivity in distinguishing RNA concentration gradients across two cancer cell lines (p < 0.0001) (Fig. 6D). Critically, the MIRACLE SlipChip system successfully detected low-abundance hsa-miR-338-3p in SKOV3 cells (36.249 ± 0.41 copies per μL), whereas RT-qPCR failed to quantify the target due to exceeding the validated cycle threshold (Ct = 38.6 ± 1.2). These results confirm the high sensitivity and reliability of the MIRACLE SlipChip for detecting low-abundance RNA in OC cell lines (Fig. S8 and S9).

This study conducted an analysis of clinically collected blood samples from a multi-stage cohort (24 pathologically confirmed OC patients: 8 stage IV, 9 stage III, 7 stages I and II; 24 healthy controls) to assess the cancer diagnostic capability of the MIRACLE SlipChip platform. Calibration curves quantified the concentration profiles of hsa_circ_0049101 and hsa-miR-338-3p, with the platform accurately identifying most positive cases (Fig. 6C). The results revealed stage-dependent expression patterns, both RNA concentrations exhibited a significant stepwise increase with cancer progression (stage IV > III > I–II), showing 2.3- to 3.7-fold higher expression in advanced-stage patients (III and IV, n = 17) compared to early-stage cases (I and II, n = 7) (p < 0.0001). Compared to RT-qPCR (Fig. 6E and F), the platform demonstrated an 18.6% improvement in positive detection rate (p < 0.01) and successfully identified low-abundance targets, such as trace circRNA in sample #9 (Ct = 39.9). Notably, the MIRACLE SlipChip quantified RNA below 38 copies per μL, surpassing RT-qPCR's effective detection range (Ct = 15.4–37.8). These findings validate the platform's accuracy for cancer diagnostics (Fig. S10–S13).

RT-qPCR quantification of hsa_circ_0049101 and hsa-miR-338-3p was performed in an identical clinical cohort, revealing a significant inverse correlation between MIRACLE SlipChip and RT-qPCR outcomes (Pearson r = −0.953 and −0.947, p < 0.0001; Fig. 6G and H). ROC analysis assessed MIRACLE SlipChip's diagnostic performance for circRNA and miRNA in all stage and early-stage samples versus conventional biomarkers (CA-125, HE4, ROMA index) (Fig. 6I and J). The platform demonstrated superior sensitivity and specificity over serum biomarkers in full-stage analyses, with RNA biomarkers achieving an AUC of 0.985 (95% CI: 0.649–0.926) for hsa_circ_0049101 and 0.964 (95% CI: 0.818–1.000) for hsa-miR-338-3p. (CA-125: 0.82–0.89). Compared with clinical serum biomarker detection methods, this approach achieved both sensitivity and specificity exceeding 90% (Table 3). For early-stage samples (I and II), the method outperformed clinical biomarkers, showing AUCs of 0.969 (95% CI: 0.649–0.926) for hsa_circ_0049101 and 0.980 (95% CI: 0.818–1.000) for hsa-miR-338-3p. Compared with clinical serum biomarker detection methods, this approach achieved both sensitivity and specificity exceeding 85% (Table 4).

Table 3 AUCs, sensitivities, and specificities of the MIRACLE SlipChip assay, CA-125 ELISA, HE4 ELISA techniques and ROMA index for all stage samples

	Biomarker (method)	AUC (95% CI)	Cutoff value	Sensitivity (%) (95% CI)	Specificity (%) (95% CI)
CI, confidence interval.
OV versus NC	Hsa_circ_0049101 (copies per μL)	0.985	35	95.8	95.8
	Hsa-miR-338-3p (copies per μL)	0.964	34.5	100	91.7
	ELISA (CA-125 (U mL^−1))	0.837	20.63	70.8	95.8
	ELISA (HE4 (pmol L^−1))	0.836	31.56	79.2	79.2
	ELISA (ROMA (%))	0833	2.16	87.5	75
	ELISA (ROMA (%))	0.864	9.58	79.2	91.7

---

Table 4 AUCs, sensitivities, and specificities of the MIRACLE SlipChip assay, CA-125 ELISA, HE4 ELISA techniques and ROMA index for early-stage samples

	Biomarker (method)	AUC (95% CI)	Cutoff value	Sensitivity (%) (95% CI)	Specificity (%) (95% CI)
CI, confidence interval.
OV versus NC	Hsa_circ_0049101 (copies per μL)	0.969	35	85.7	100
	Hsa-miR-338-3p (copies per μL)	0.980	68.5	85.7	100
	ELISA (CA-125 (U mL^−1))	0.684	20.4	42.9	100
	ELISA (HE4 (pmol L^−1))	0.776	17.25	100	71.4
	ELISA (ROMA (%))	0.939	1.895	85.7	100
	ELISA (ROMA (%))	0.918	3.445	85.7	85.7

---

These results confirm the platform's high sensitivity, specificity, and robustness for RNA analysis, highlighting its potential for liquid biopsy-based early cancer detection. While the current validation with 48 clinical samples (including 7 early-stage cases) demonstrates feasibility, further multi-center studies with larger cohorts (particularly enriched in early-stage patients) are required to confirm generalizability. This study establishes an innovative solution for OC liquid biopsy, bridging methodological advancement to clinical translation.

4. Conclusion

In this work, we established MIRACLE (Modular Isothermal RCA-Activated Cas12a Loop-Enhanced amplification), a reverse-transcription-free biosensing platform that integrates programmable molecular circuits with a tunable SlipChip microfluidic system for ultrasensitive quantification of circRNA and miRNA biomarkers. The system leverages three foundational innovations: (i) target-agnostic circDNA scaffolds transform RNA inputs into uniform Cas12a-triggering templates; (ii) BstNI-directed circRNA back-splice junction identification, obviating RT steps and linear RNA artifacts; (iii) CRISPR-coupled microfluidic amplification with exponential signal enhancement through cis/trans-cleavage feedback. The MIRACLE SlipChip platform achieves single-copy sensitivity (circRNA: 0.125 copies per μL; miRNA: 0.326 copies per μL) within 35 min, representing a 10-fold improvement over conventional CRISPR-based techniques. This performance outperforms gold-standard RT-qPCR and microarrays in critical metrics, such as speed, specificity, and cost-efficiency. Clinical validation using ovarian cancer cell lines and blood specimens demonstrated robust detection of hsa_circ_0049101 and hsa-miR-338-3p across staging, underscoring its potential for liquid biopsy application. This platform's modular design with interchangeable components enables broad adaptability to other disease related RNAs. This breakthrough offers an efficient solution for early precision screening in resource-limited settings, bridging methodological innovation and clinical scalability in CRISPR diagnostics.

Author contributions

Lingxi Tian: writing – review & editing, writing – original draft, visualization, validation, methodology, conceptualization. Yan Gao: writing – original draft, investigation. Yang Lu: visualization, methodology, investigation. Feng Xu: investigation. ZiRui Feng: formal analysis, investigation, visualization. Lihan Zi: investigation. Zaian Deng: writing – review & editing, supervision. Jun Yang: writing – review & editing, supervision, project administration, funding acquisition.

Conflicts of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

Supplementary information is available: Detailed reagents and methodologies, including cell culture conditions, procedures for RNA in vitro transcription, as well as the design and fabrication process of the microfluidic chip. Fig. S1–S16, which provide additional results supporting the main findings. These include mechanistic studies of Cas12a cleavage activity, detection range of the MIRACLE method, amplification efficiency experiments, specificity assessments, condition optimization data, and original image data. Tables S1–S3, listing oligonucleotide sequences used in the study and clinical characteristics of the patient cohort. See DOI: https://doi.org/10.1039/D5LC00585J.

The data supporting this article have been included as part of the SI.

Acknowledgements

This work was supported by the Fundamental Research Funds for the Central Universities, the Dalian Science and Technology Innovation Fund (No. 2022JJ12SN049), and the Dalian Life and Health Field Guidance Plan (No. 2024ZDJH01PT012).

References

1. B. F. Han, R. S. Zheng and H. M. Zeng, et al., Cancer incidence and mortality in China, J. Natl. Cancer Cent., 2024, 4(1), 47–53.
2. G. C. Jayson, E. C. Kohn, H. C. Kitchener and J. A. Ledermann, Ovarian cancer, Lancet, 2014, 384(9951), 1376–1388.
3. M. McMullen, K. Karakasis, R. Rottapel and A. M. Oza, Advances in ovarian cancer, from biology to treatment, Nat. Cancer, 2021, 2(1), 6–8.
4. V. Dochez, H. Caillon, E. Vaucel, J. Dimet, N. Winer and G. Ducarme, Biomarkers and algorithms for diagnosis of ovarian cancer: CA125, HE4, RMI and ROMA, a review, J. Ovarian Res., 2019, 12(28), 2–9.
5. Z. Foruzandeh, F. Zeinali-Sehrig and K. Nejati, et al., CircRNAs as potent biomarkers in ovarian cancer: a systematic scoping review, Cell. Mol. Biol. Lett., 2021, 26, 41.
6. Y. Wu, Y. Zhang and X. H. Zhang, et al., Research advances for exosomal miRNAs detection in biosensing: From the massive study to the individual study, Biosens. Bioelectron., 2021, 177(1), 112962.
7. S. Memczak, M. Jens and A. Elefsinioti, et al., Circular RNAs are a large class of animal RNAs with regulatory potency, Nature, 2013, 495(7441), 333–338.
8. Y. Li, Q. Zheng and C. Bao, et al., Circular RNA is enriched and stable in exosomes: A promising biomarker for cancer diagnosis, Cell Res., 2021, 25(8), 981–984.
9. H. Jeon, S. M. Seo and T. W. Kim, et al., Circulating Exosomal miR-1290 for Diagnosis of Epithelial Ovarian Cancer, Curr. Issues Mol. Biol., 2022, 44(1), 288–300.
10. S. Cai, T. Pataillot-Meakin and A. Shibakawa, et al., Single-molecule amplification-free multiplexed detection of circulating microRNA cancer biomarkers from serum, Nat. Commun., 2021, 12, 3515.
11. T. Xu, M. Wang and L. Jiang, et al., CircRNAs in anticancer drug resistance: Recent advances and future potential, Mol. Cancer, 2020, 19(1), 127.
12. L. S. Kristensen, T. B. Hansen, M. T. Venø and J. Kjems, Circular RNAs in cancer: Opportunities and challenges in the field, Oncogene, 2018, 37(5), 555–565.
13. Y. Li, Q. Zheng and C. Bao, et al., Circular RNA is enriched and stable in exosomes: A promising biomarker for cancer diagnosis, Cell Res., 2021, 25(8), 981–984.
14. R. B. Ankur, P. Ritu and A. Manu, et al., Northern Blotting Technique for Detection and Expression Analysis of mRNAs and Small RNAs, Methods Mol. Biol., 2021, 2170, 155–183.
15. J. Zhang and F. Zhao, Circular RNA discovery with emerging sequencing and deep learning technologies, Nat. Genet., 2025, 57, 1089–1102.
16. M. Dellett and D. A. Simpson, Considerations for optimization of microRNA PCR assays for molecular diagnosis, Expert Rev. Mol. Diagn., 2016, 16, 407–414.
17. C. Pritchard, H. Cheng and M. Tewari, MicroRNA profiling: approaches and considerations, Nat. Rev. Genet., 2012, 13, 358–369.
18. R. T. Fuchs, Z. Sun, F. Zhuang and G. B. Robb, Bias in ligation-based small RNA sequencing library construction is determined by adaptor and RNA structure, PLoS One, 2015, 10(5), e0126049.
19. A. F. Zhaleh, X. L. Zhu and H. C. Wang, et al., Embarking on a journey through Micro-RNA and Circular-RNA detection methods, TrAC, Trends Anal. Chem., 2024, 181, 118035.
20. J. P. K. Bravo, T. Hallmark and B. Naegle, et al., RNA targeting unleashes indiscriminate nuclease activity of CRISPR-Cas12a2, Nature, 2023, 613(7944), 582–587.
21. D. C. Swarts and M. Jinek, Mechanistic Insights into the cis- and trans-Acting DNase Activities of Cas12a, Molecular Cell, 2019, 73(3), 589–600.
22. J. S. Chen, E. B. Ma and L. B. Harrington, et al., CRISPR-Cas12a target binding unleashes indiscriminate single-stranded DNase activity, Science, 2018, 360(6387), 436–439.
23. X. X. Ke, A. J. Liang and C. X. Chen, et al., A one-pot CRISPR-RCA strategy for ultrasensitive and specific detection of circRNA, Anal. Methods, 2024, 16, 3256–3262.
24. I. A. Murray, S. K. Stickel and R. J. Roberts, Sequence-specific cleavage of RNA by Type II restriction enzymes, Nucleic Acids Res., 2010, 38, 8257–8268.
25. G. Jake Carter, L. O. Iturbe and J.-L. H. A. Duprey, et al., Ultrarapid detection of SARS-CoV-2 RNA using a reverse transcription-free exponential amplification reaction, RTF-EXPAR, Proc. Natl. Acad. Sci. U. S. A., 2021, 118(35), e2100347118.
26. H. Yan, Y. Wen and Z. Tian, et al., A one-pot isothermal Cas12-based assay for the sensitive detection of microRNAs, Nat. Biomed. Eng., 2023, 7, 1583–1601.
27. G. Nallur, C. Luo and L. Fang, et al., Signal amplification by rolling circle amplification on DNA microarrays, Nucleic Acids Res., 2001, 29(23), E118.
28. J. Baner, M. Nilsson and M. Mendel-Hartvig, et al., Signal amplification of padlock probes by rolling circle replication, Nucleic Acids Res., 1998, 26(22), 5073–5078.
29. D. Fredrik, J. Banér and M. Gullberg, et al., Circle-to-circle amplification for precise and sensitive DNA analysis, Proc. Natl. Acad. Sci. U. S. A., 2004, 101(13), 4548–4553.
30. S. M. Wang, K. Zhang and S. Y. Tan, et al., Circular RNAs in body fluids as cancer biomarkers: the new frontier of liquid biopsies, Mol. Cancer, 2021, 20(1), 13.
31. Q. Y. Zhao, J. Y. Liu and H. Deng, et al., Targeting mitochondria-located circRNA SCAR alleviates NASH via reducing mROS output, Cell, 2020, 183(1), 76–93.
32. Z. D. Li, Y. L. Li and J. Lu, et al., Recent research progress of circular RNAs in hepatocellular carcinoma, Front. Oncol., 2024, 13, 1192386.
33. J. E. Kreutz, T. Munson and T. Huynh, et al., Theoretical design and analysis of multivolume digital assays with wide dynamic range validated experimentally with microfluidic digital PCR, Anal. Chem., 2011, 83(21), 8158–8168.
34. F. Shen, B. Sun and J. E. Kreutz, et al., Multiplexed Quantification of Nucleic Acids with Large Dynamic Range Using Multivolume Digital RT-PCR on a Rotational SlipChip Tested with HIV and Hepatitis C Viral Load, J. Am. Chem. Soc., 2011, 133, 1770–17712.
35. S. Jonathan, O. O. Gootenberg and M. J. Abudayyeh, et al., Multiplexed and portable nucleic acid detection platform with Cas13, Cas12a, and Csm6, Science, 2018, 360, 439–444.
36. G. Hu, Z. J. Wei and J. L. Guo, et al., A Single-Tube Two-Step MIRA-CRISPR/Cas12b Assay for the Rapid Detection of Mpox Virus, Viruses, 2025, 17(6), 841.
37. G. Y. Tang, Z. L. Zhang and W. Tan, et al., RT-RPA-Cas12a-based assay facilitates the discrimination of SARS-CoV-2 variants of concern, Sens. Actuators, B, 2023, 381(15), 133433.
38. K. Wang, H. F. Yin and S. Li, et al., Quantitative detection of circular RNA and microRNA at point-of-care using droplet digital CRISPR/Cas13a platform, Biosens. Bioelectron., 2025, 267(1), 116825.
39. H. Y. Jia, H. L. Zhao and T. Wang, et al., A programmable and sensitive CRISPR/Cas12a-based MicroRNA detection platform combined with hybridization chain reaction, Biosens. Bioelectron., 2022, 211(1), 114382.
40. Y. X. Gong, H. Y. Zeng and S. S. Gao, Application of CRISPR/Cas12a in miRNA-155 detection: A novel homogeneous electrochemiluminescence biosensor, Anal. Chim. Acta, 2024, 1316(8), 342843.
41. L. J. Sha, J. H. Yao and S. Yang, et al., Collaborative CRISPR-Cas System-Enabled Detection of Circulating Circular RNA for Reliable Monitoring of Acute Myocardial Infarction, Small, 2024, 20(45), 2402895.
42. B. Jiang, J. Zhang and X. Sun, et al., Circulating exosomal hsa_circRNA_0039480 is highly expressed in gestational diabetes mellitus and may be served as a biomarker for early diagnosis of GDM, J. Transl. Med., 2022, 20(5), 2–13.
43. L. X. Tian, Y. Gao and L. H. Zi, et al., Dual Cas12a and Multiplex crRNA CRISPR strategy ultrasensitive detection novel circRNA biomarker for the diagnosis of ovarian cancer, BMC Cancer, 2025, 25, 695.
44. Y. Alghamian, C. Soukkarieh, A. Aljapawe and H. Murad, Exploring miRNA profile associated with cisplatin resistance in ovarian cancer cells, Biochem. Biophys. Rep., 2025, 41, 101906.

---