How physical blocking controls CRISPR/Cas12a's cutting speed using split activator extensions

The DNA oligonucleotides and RNA used in this study were synthesized and high pressure (or high performance) liquid chromatography-purified by Shanghai Sangon Biotech Co., Ltd. (Shanghai, China). Detailed sequences and modification information are listed in. Lba Cas12a (Cpf1, 100 μM) and 10× rCutSmart buffer (50 mM potassium acetate, 20 mM Tris-acetate, 10 mM magnesium acetate, 100 μg/ml recombinant albumin, pH 7.9) were purchased from New England Biolabs (Beijing, China). Tris-EDTA (TE) buffer [10 mM Tris–HCl, 1 mM ethylenediaminetetraacetic acid (EDTA), pH 8.0], Diethyl pyrocarbonate (DEPC)-treated water (RNase-free), dithiothreitol (DTT), magnesium chloride hexahydrate (MgCl₂·6H₂O), and sodium chloride (NaCl) were purchased from Shanghai Sangon Biotech Co., Ltd. (Shanghai, China). Supplementary Table S1

Fluorescence signal monitoring was performed using a real-time fluorescence quantitative polymerase chain reaction (PCR) system FQD-96A (BIOER, China). The reaction temperature was set at 37°C, the green fluorescence channel was selected, and the gain was set to the default value. Oligonucleotide annealing reactions were also performed on this PCR instrument.

DNA oligonucleotides were dissolved in TE buffer (10 mM Tris–HCl, 1 mM EDTA, pH 8.0), and RNA oligonucleotides were dissolved in DEPC water. DNA complexes were assembled in TE/Mg²⁺ buffer [10 mM Tris–HCl, 1 mM EDTA, 12.5 mM MgCl₂, pH 8.0] by mixing corresponding complementary single strands in equimolar concentrations. The annealing process was performed on a PCR instrument: heated at 95°C for 5 min, and then ramped down to 20°C at a rate of 0.1°C/s. After annealing, the complexes were stored at 4°C for later use.

To evaluate how different split-activator architectures regulate Cas12a trans-cleavage activity, we examined the performance of various ssDNA and dsDNA split activators both individually and in their assembled mixed states. Each reaction (20 μl) contained 20 nM ribonucleoprotein (RNP), 10 nM of the corresponding split-activator components, 250 nM reporter, 1× rCutSmart buffer, and 1 mM DTT, with the remaining volume supplemented using TE/Mg²⁺ buffer. Reactions were carried out at 37°C, and fluorescence kinetics were recorded in real time using a quantitative PCR fluorescence reader.

To investigate kinetic matching within the cascaded reaction, we constructed and compared three activation strategies: canonical activation, conventional split activation, and steric-regulated split activation. A typical one-pot reaction (20 μl) contained 100 nM three-strand substrate complex, 100 nM fuel strand, 10 nM miRNA-21, 20 nM RNP, 250 nM reporter, 1× rCutSmart buffer, and 1 mM DTT. For strategies involving split activation, 100 nM of the fixed-end split strand was preincubated with RNP to form the initial complex, while the output strand generated by the EDC cycle served as the complementary activating component.

After identifying the optimal steric-regulated split-activator design, we further evaluated the analytical performance of the optimized one-pot system. The evaluation mixture (20 μl) consisted of 100 nM S0’–S1–S2’, 100 nM F’, 80 nM RNP, 100 nM Pd-ss11_DE3, 250 nM reporter, 1× rCutSmart buffer, and 1 mM DTT. By introducing various concentrations of miRNA-21 or interfering sequences and monitoring fluorescence at 37°C in real time, we assessed the sensitivity, linear range, and specificity of the cascaded system.

Fluorescence data were normalized to quantify the regulatory effects of different split-activator designs on Cas12a activity. The relative cleavage efficiency was calculated using the equation: % cleaved = F/F_PC × 100%. Here, F denoted the fluorescence intensity of the experimental group, and F_PC represented the plateau fluorescence of the Positive Control (PC), defined as the reaction containing the canonical full-length TS that yielded maximal trans-cleavage activity. The Negative Control represented the baseline signal and was obtained from reactions containing all components except the activator strands. Because different regulatory strategies originated from distinct initial structural states, the Control (C) condition was defined operationally for each strategy as the unmodified baseline design prior to introducing steric regulatory elements. Depending on the specific strategy being evaluated, the C condition corresponded to a native ssDNA split pair, a dsDNA-anchored construct, or a minimally extended variant. In each case, the C condition served as the internal reference for evaluating the quantitative impact of subsequent structural modifications. By normalizing all experimental signals to the canonical full-length activator (PC) and referencing each strategy against its internal baseline (C), this framework enabled consistent comparison of relative activation efficiencies across distinct steric-regulated designs.

All quantitative data were obtained from at least three independent experimental replicates and are presented as mean ± standard deviation (SD), unless otherwise noted. Data preprocessing, curve fitting, kinetic parameter extraction, and signal-to-background calculations were performed using Origin 2024 (OriginLab, USA). Graphs and numerical summaries were generated using Origin 2024 and GraphPad Prism 9 (GraphPad Software, USA).

To construct a programmable and ultra-low-background Cas12a regulatory platform, the primary task is to identify a set of split activators that remain silent in isolation but efficiently activate Cas12a upon assembly. Although a single short ssDNA is typically insufficient to induce the conformational changes required for trans-cleavage, the correct assembly of two complementary split strands guided by crRNA can effectively restore enzymatic activity (Fig. 1A). However, the choice of the split site fundamentally dictates the binding energy of the fragments, the thermodynamic stability of the assembly, and the risk of potential single-strand leakage. To this end, we systematically designed nine sets of ssDNA combinations with varying split topologies based on the complementary sequence of the crRNA spacer (designated as Pd-ss x and Pp-ss y, Fig. 1B) and screened their activation kinetics. As shown in Fig. 1C–K, the split site exerted a decisive impact on system performance. In combinations with longer Pp fragments (Groups 1–3), significant single-strand leakage was observed. Notably, in Group 1 (Fig. 1C), the addition of single-stranded Pp-ss14 alone induced a marked rise in fluorescence signal, indicating that the overly long single-stranded fragment is capable of independently inducing significant background cleavage activity. Conversely, when the Pd fragment was longer while the Pp fragment was too short (Groups 7 and 8), the single-strand leakage was effectively suppressed, but the overall activation efficiency upon mixing was significantly compromised (Fig. 1I–J). For Group 9 (Fig. 1K), although its mixed activation efficiency recovered to a relatively high level, the overly long Pd fragment (14 nt) led to the reappearance of single-strand background signal. In contrast, Groups 4, 5, and 6 achieved a favorable balance between activation activity and background noise (Fig. 1F–H).

To identify the optimal candidate among these preferred groups, performance was quantitatively evaluated using the SBR (Fig. 1M), calculated by dividing the peak fluorescence of the mixed assembly by the maximum background leakage generated by either single-stranded component. This evaluation strategy benchmarks against the system’s most severe potential leakage, reflecting true activation efficacy. Given the marginal differences among the three groups within the standard reaction time, we further performed a 180-min long-term kinetic validation (Supplementary Fig. S1). The results showed that the single-stranded components of Groups 4 and 5 exhibited slight signal drift over time, suggesting a risk of metastable binding; only Group 6 maintained its single-stranded components at an extremely low baseline level throughout the monitoring window, achieving a significantly higher ratio compared to other combinations. Based on this rigorous screening, we established Pd-ss11 and Pp-ss9 as the optimal split-activator combination. This combination nearly eliminates basal leakage mediated by single components while retaining high assembly-dependent activation capacity, providing an ideal low-background structural foundation for the subsequent introduction of programmable steric regulation via terminal extensions.

To elucidate the molecular basis underlying kinetic regulation, we first employed flexible ssDNA as the anchoring scaffold of the split-activator system, introducing terminal extensions on the partner strand to impose programmable steric hindrance. Using Pd-ss11 and Pp-ss9 as the core split modules, we fixed one strand and progressively extended the 5′ end, 3′ end, or both ends (3–12 nt) of the other strand to systematically evaluate how distinct spatial configurations influence Cas12a activation kinetics. For clarity, extended strands are designated by appending 5′E (5′-end extension), 3′E (3′-end extension), or DE (dual-end extension) followed by length to the core name, with ss and ds denoting single-stranded and double-stranded forms, respectively.

We first examined the configuration in which Pp-ss9 was fixed while Pd-ss11 was varied (Fig. 2A). Real-time fluorescence kinetic revealed pronounced directional sensitivity (Supplementary Fig. S2). 3′-end extensions markedly suppressed both the activation rate and signal amplitude in a length-dependent manner, whereas 5′-end extensions had only minimal impact (Fig. 2B and C). Dual-end extensions produced additive inhibitory effects, substantially stronger than those of single-end extensions, indicating that the 3′ terminus of Pd-ss11 represents a sterically sensitive hotspot. When these flexible ssDNA extensions were replaced with rigid dsDNA extensions (Fig. 2D), steric hindrance effects were further amplified. Kinetic monitoring (Supplementary Fig. S3) showed that 3′ dsDNA extensions sharply reduced Cas12a activity, and dual dsDNA extensions nearly abolished activation, demonstrating the physical additivity of dual-end steric obstruction (Fig. 2E and F).

We next evaluated the complementary anchoring configuration, in which Pd-ss11 was fixed and Pp-ss9 was extended with terminal ssDNA additions (Fig. 2G). Under ssDNA extensions, the system exhibited a mirror-like regulatory pattern (Supplementary Fig. S4). 5′-end extensions acted as the dominant inhibitory factor, causing pronounced delays in activation, whereas 3′-end extensions did not suppress activity and instead maintained or increased it (Fig. 2H and I), suggesting that this terminus may adopt local conformations favorable for R-loop formation. Nonetheless, dual-end extensions (Pp-ss9_DE) still produced marked inhibition due to the presence of the sensitive 5′-end steric site, demonstrating the robustness of dual-end strategies even when one terminus is intrinsically tolerant. When ssDNA extensions were replaced with rigid dsDNA extensions (Fig. 2J), steric suppression was further strengthened (Supplementary Fig. S5). 5′ dsDNA extensions significantly delayed activation, while dual dsDNA extensions exerted the strongest cumulative blocking effect, reducing the fluorescence signal and overriding the noninhibitory nature of the 3′ terminus (Fig. 2K and L). To assess whether these steric effects introduced by the duplex scaffold were influenced by PAM specificity, we evaluated a representative combination (fixed Pd-ss11 with Pp-ds9_DE6) across four PAM variants (canonical TTTA and three noncanonical sequences; Supplementary Fig. S6). Although PAM identity shaped the intrinsic activation rate and signal magnitude, the fundamental steric regulatory trend remained unchanged, indicating that steric modulation functions as a physical control layer independent of sequence-specific recognition.

Together, these comparative analyses establish three fundamental principles governing steric regulation within ssDNA-anchored systems: (i) directional asymmetry and length dependence, with Pd-3′ and Pp-5′ identified as steric hotspots; (ii) dual-end extensions generate additive steric barriers, enabling effective inhibition even when one terminus is nonsensitive; and (iii) rigid structures amplify steric obstruction, providing stronger kinetic barriers than flexible ssDNA. These findings demonstrate that Cas12a activation kinetics can be precisely regulated by rationally designing the geometry, physical rigidity, and extension length of split-activator termini.

Having established the principles of steric regulation in ssDNA scaffolds, we next investigated how steric effects evolve when the anchoring strand is replaced by a more rigid dsDNA. Short duplexes (9 or 11 bp) exhibit insufficient thermal stability at 37°C, making unextended Pd-ds11 or Pp-ds9 unable to maintain rigid conformations. To obtain stable duplex anchoring points without introducing background activation, we introduced a necessary 6-nt extension at the nonsensitive terminus of the fixed strand, as guided by the principles derived above, generating Pd-ds11_5′E6 and Pp-ds9_3′E6. NUPACK simulations confirmed that both designs formed highly stable duplexes under reaction conditions (), ensuring that observed kinetic differences predominantly arose from the intended steric manipulations rather than scaffold dissociation. Supplementary Fig. S7

We first anchored the rigid Pp-ds9_3′E6 scaffold and introduced extensions on Pd-ss11 (Fig. 3A). Compared with the ssDNA-anchored system, this configuration exhibited markedly enhanced steric sensitivity (Supplementary Fig. S8). As shown in Fig. 3B and C, 3′-end ssDNA extensions on Pd-ss11 strongly suppressed Cas12a activation in a sharply length-dependent manner, and dual-end extensions produced additive inhibitory effects. When Pd ssDNA extensions were replaced with rigid dsDNA extensions (Fig. 3D and Supplementary Fig. S9), suppression became even stronger. 3′ dsDNA extensions nearly extinguished fluorescence signals, whereas dual-end dsDNA extensions served as an almost complete physical “blocker,” reducing activity to background levels across all tested lengths (Fig. 3E and F), demonstrating synergistic steric exclusion contributed by the rigid scaffold and rigid extensions.

We next analyzed the complementary configuration, where Pd-ds11_5′E6 was fixed and Pp-ss9 was extended (Fig. 3G). In this arrangement, Pp-ss9_5′ extensions remained the dominant inhibitory contributor (Fig. 3H and I, and Supplementary Fig. S10). Conversely, Pp-ss9_3′ extensions did not exhibit inhibition across the examined lengths and instead increased activity, again indicating high conformational tolerance at this terminus. As a result, Pp-ss9_DE displayed intermediate inhibition, with the noninhibitory 3′ end partially buffering the 5′ steric effect. However, when Pp ssDNA extensions were replaced with rigid dsDNA (Fig. 3J and Supplementary Fig. S11), steric suppression regained full dominance. 5′ dsDNA extensions markedly diminished activity, and dual-end dsDNA extensions effectively overcame the inherent 3′-end tolerance, yielding near-complete inhibition of Cas12a activation (Fig. 3K and L). To assess the generality of rigidity-amplified steric regulation, we examined the fully rigid combination (Pd-ds11_5′E6 with Pp-ds9_DE6) across different PAM variants (Supplementary Fig. S12). PAM identity continued to modulate activation strength; however, relative to ssDNA systems, introducing a second rigid anchor substantially reduced fluorescence across all PAM contexts, indicating that duplex anchoring restricts molecular degrees of freedom to generate PAM-independent global inhibition.

Based on these systematic investigations, we established the following universal quantitative principles of steric regulation: (i) strong directional specificity, with Pd-3′ and Pp-5′ serving as key steric hotspots; (ii) regulation strength can be continuously tuned by extension rigidity and length, with rigid duplex scaffolds markedly amplifying steric effects and lowering inhibitory thresholds; (iii) conformational tolerance at specific termini can be effectively overcome by rigid duplex extensions; and (iv) dual rigid-end extensions integrate steric constraints to produce a robust “shutdown” mode, suppressing Cas12a activation to near-background levels. These principles constitute a programmable “structural rule” framework that provides clear design guidance for engineering high-performance Cas12a systems with rationally tailored steric regulation.

Nucleic acids, with their high programmability, predictable thermodynamic properties, and low synthesis costs, have become ideal components for molecular nanotechnology [36, 37]. These characteristics have driven the development of various self-assembled structures and dynamic reaction networks, enabling precise spatiotemporal control through mechanisms such as strand displacement and extension [38, 39]. As a typical nonenzymatic amplification strategy, the EDC can achieve efficient signal conversion from TSs to output strands through strand-displacement reactions, making it an ideal choice for molecular diagnostics [40, 41]. Therefore, we further explored whether it is possible to achieve precise control of CRISPR/Cas programmable activation, without external stimuli or additional modifications, solely relying on the inherent programmability of nucleic acids. This would allow it to be effectively isolated in the temporal domain from the upstream reaction network, thereby establishing an efficient, low-background one-pot cascaded system.

However, when integrating the EDC and CRISPR/Cas12a systems into a single reaction vessel for a one-pot cascade, a pronounced kinetic mismatch arises: the trans-cleavage rate of Cas12a is far faster than the strand-displacement amplification rate of the EDC, leading to substantial signal crosstalk. To clearly illustrate this issue and demonstrate the solution, we systematically compared three Cas12a activation strategies (Fig. 4A). In the canonical activation strategy (Fig. 4A and I), a full-length single-stranded activator perfectly matching the EDC output is directly employed. Owing to the absence of kinetic control, trace amounts of nonspecific intermediates are sufficient to instantly trigger burst-phase Cas12a cleavage, resulting in severe background leakage and premature saturation of the detection signal before the EDC has accumulated meaningful output (Fig. 4B). In the regular split activation strategy (Fig. 4A-II), dividing the activator into two short strands reduces basal activity to some extent; however, the spontaneous hybridization kinetics of the split strands remain poorly matched to the slower EDC amplification process, yielding appreciable background signals nonetheless (Fig. 4C).

To fundamentally overcome this bottleneck, we developed a steric-regulated split activation strategy, guided by the steric modulation principles established earlier. The key concept is to introduce programmable terminal extensions onto the split activators, thereby installing a spatial barrier between the activator and the Cas12a complex and establishing a defined kinetic activation threshold for Cas12a. As shown in Fig. 4A-III, this strategy effectively reshapes the cascade reaction kinetics. During the early reaction stage, the steric obstruction introduced by the terminal extensions suppresses low-concentration nonspecific assembly of the activator–Cas12a complex, enabling the EDC to accumulate output without premature Cas12a activation. Only when the EDC cycle releases sufficiently high concentrations of output strands capable of overcoming the steric barrier does Cas12a become specifically activated. As demonstrated in Fig. 4D and E, this strategy markedly decreases background signals and enhances the signal-to-noise ratio (S/N, where S and N correspond to fluorescence intensities in the presence and absence of target, respectively), achieving near-leakage-free cascade detection and effectively resolving the kinetic mismatch. In Supplementary Figs S13–S17, we further screened combinations of fixed strands and extension designs. The results showed that the configuration employing fixed Pd-ss11_DE3 together with S1’ as the activator strand achieved the best performance, efficiently suppressing background while retaining high responsiveness to target-triggered signal generation. This design provides a new conceptual framework for the deep integration of programmable nucleic acid systems with CRISPR-based detection platforms.

Based on the above findings, to further verify the reliability and practical value of the constructed cascaded system, we systematically evaluated its key performance metrics, including sensitivity and specificity. The detection principle is shown in Fig. 5A. The reaction mixture contains a pre-hybridized three-strand substrate complex (S0’–S1–S2’), a fuel strand (F’), and a Cas12a RNP complex preassembled with the sterically regulated split activator (Pd-ss11_DE3). In the absence of target, the terminal extension on the Pd strand imposes steric hindrance that kinetically locks Cas12a in an inactive state, thereby effectively suppressing background leakage. Upon the introduction of miRNA-21, the target initiates a toehold-mediated strand-displacement reaction with S0’, releasing S1, and exposing the previously concealed toehold on S0’. The fuel strand F’ subsequently drives the second displacement step, liberating both the recyclable TS and the output strand S2’. Accumulated S2’ functions as a critical unlocking species that cooperatively assembles with Pd-ss11_DE3 and overcomes the steric barrier, specifically activating Cas12a trans-cleavage and generating amplified fluorescence.

To achieve optimal analytical performance, we optimized the fuel strand sequence and RNP concentration (Supplementary Figs S18–S19). The results indicated that the F’4 fuel strand combined with 80 nM RNP yielded the highest S/N and was thus selected for subsequent assays. Under these optimized conditions, we next evaluated the system’s quantitative performance across a series of miRNA-21 concentrations. As shown in Fig. 5B, the fluorescence intensity increased progressively as the miRNA-21 concentration rose from 10 pM to 5 nM. Within the range of 10 pM to 1 nM, fluorescence intensity exhibited a strong linear correlation with target concentration (Fig. 5C), described by the regression equation y = 13.8411x + 2262.38, with a correlation coefficient of R² = 0.9983. The limit of detection calculated based on the 3σ criterion was 1.24 pM [42, 43].

To evaluate the specificity of the strategy, we tested various interfering miRNAs (including miRNA-155, miRNA-373, miRNA-141, let-7a) as well as single-base mismatch (SM-miRNA-21) and double-base mismatch (DM-miRNA-21) sequences. As shown in Fig. 5D, only miRNA-21 induced a significant fluorescence increase, while the signals from other miRNAs and mismatch sequences were comparable to the blank control group. This indicates that the cascaded system has high specificity for the target miRNA-21.

Collectively, the sterically regulated one-pot cascade system exhibits picomolar-level sensitivity, a broad linear dynamic range, and single-nucleotide specificity, highlighting its potential application value in early disease biomarker detection and complex sample analysis, and facilitating the deep integration of programmable nucleic acid reaction networks with CRISPR technology.

All data supporting the findings of this study are available within the article and its supplementary information or will be made available from the authors upon request.