DNA computing function switching by programming base stacking interactions with minimal molecular architecture changes

Zhang, Yongpeng; Li, Bozhao; Liu, Xuan; Zheng, Xuedong; Liu, Shi; Nie, Guangjun; Yang, Jing; Ke, Yonggang; Li, Suping; Zhang, Cheng

doi:10.1038/s41467-025-64406-x

Download PDF

Article
Open access
Published: 23 October 2025

DNA computing function switching by programming base stacking interactions with minimal molecular architecture changes

Nature Communications volume 16, Article number: 9369 (2025) Cite this article

7914 Accesses
4 Citations
Metrics details

Subjects

Abstract

In biological systems, molecular network functionalities are usually switched in a flexible, facile, and programmable manner. Mimicking this, substantial studies are directed towards developing synthetic DNA networks that exhibit similar function-switching capabilities, though often hindered by extensive molecular architecture changes and stringent condition controls, which result in a time-consuming and labor-intensive process. Here, we develop a base stacking-mediated allostery strategy to manipulate the DNA computing function switching with minimal molecular architecture changes, usually as few as 1-2 nucleotide changes. We implement up to 20 distinct logic function switching within DNAzyme networks. We also validate our function switching platform to implement totally 84 kinds of gene regulation patterns in cancer cell lines, demonstrating its utility in RNA sensing and green fluorescent protein regulation. This strategy offers a simplified alternative approach to enrich DNA regulations, with potential applications in DNA computing and bioengineering.

Introduction

Intracellular processes are intricately regulated by molecular networks, in which diverse biomolecules can alter their functions in response to a myriad of signaling cues^1,2. The molecular function switching widely exists in various natural cellular systems, which usually involves aspects of regulating structural variations and molecular interactions^3,4,5,6. For example, the resistance functions of exoribonuclease-resistant RNAs (xrRNA) can be precisely switched just by simply varying the loop lengths of an RNA pseudoknot from 2 to 7 bp to switch Mg²⁺ dependency, showing the inherent adaptability of biomolecular systems⁷. Recognizing the significance of simple and flexible function regulations in natural systems, considerable interests have grown around the constructions of synthetic molecular networks capable of the function switching^8,9,10,11,12.

Recently, DNA computing systems have provided a programmable platform for the rational designs of synthetic DNA networks to execute diverse functions, such as DNA logic gates and nanorobots^{13,14,15,16,17,18,19,20}. Accordingly, substantial efforts are directed to develop versatile DNA computing systems with a function-switching ability^21,22, which are usually regulated by various physicochemical triggers (e.g., light and pH) and integrated DNA strand displacement reactions (SDR)^{23,24,25,26,27,28,29,30,31,32}. However, the current DNA circuit function switching strategies, often require extensive molecular architecture changes and elaborate control mechanisms, complicating the construction of efficient and scalable DNA computing systems^33,34,35. Specifically, due to a limited orthogonality in DNA circuit design, complex molecular architecture relationships intertwine among DNA gate components, thus resulting in a time-consuming and labor-intensive design process during function switching^{20,22,31,36,37}. Even in an elegantly designed general SDR-based DNA computing system, it requires ~40% changes of the DNA architectures when switching the function from AND to OR³⁸. Therefore, it is desired to develop a simple and flexible approach to perform DNA computing function switching, while with less changes of the molecular architectures.

Our study introduces a base Stacking-Mediated Allostery (SMALL) strategy that simplifies the DNA computing function switching scheme by programming stacking interactions, with minimal molecular architecture changes (usually as few as 1-2 nucleotides) (Fig. 1a–e). Utilizing orthogonal interactions between stacked bases, this allosteric approach allows for the remote activation of split DNAzyme functions (much as the buckle-up action of a safety belt), effectively reducing the need of molecular architecture changes (Fig. 1a–d and Supplementary Fig. 1). We implemented up to 20 distinct logic function switching based on SMALL strategy. These logic function switching were integrated into DNAzyme networks to implement basic, multi-input, and cascaded logic operations, achieving molecular architecture change ratios that span from 0.2% to 1.5% (Fig. 1e). We also applied SMALL strategy to totally 84 kinds of cellular gene regulation patterns (MCF-7 breast cancer cell and HeLa cervical cancer cell lines), illustrating its practical applications in real-world biological systems. The simplicity and effectiveness of our design provide an important pathway for advancing the field of DNA computing, molecular diagnostics and gene engineering.

**Fig. 1: Schematic illustrations of the SMALL function switching networks.**

Results

Basic SMALL regulated YES gate

In the basic SMALL YES gate module, the SMALL-mediated DNAzyme cleavage reaction was allosterically activated by introducing trigger D to produce stacking interaction 1 during the end’s proximity between DNA strands D and B (Fig. 2a). The basic module consists of three components: DNAzyme-1 (B/C, a basic receptor), trigger D and substrate A (containing a central ribonucleotide as the cleavage site) (Fig. 2a). Initially, DNAzyme-1 bound to substrate A via a 25 nt binding region to form complex (B/C/A), a metastable DNAzyme cleavage state (Fig. 2a). The length of the binding region between DNA strand B and substrate A was designed as 3 bp to ensure the activity of DNAzyme-1 maintained in the inactive OFF state. Upon introducing trigger D to bind with substrate A, the ends proximity of DNA strands D and B facilitated the generation of stacking interaction 1 (Fig. 2a). Assisted by stacking interaction 1, DNAzyme-1 was allosterically activated to perform the cleavage reaction (Fig. 2b).

In the YES gate configuration, DNAzyme-1 was initially designed to be inactive but could reach the activation threshold with assistance of the stacking interaction. The activity state of DNAzyme-1 was modulated by programming the stacking interactions at its left side (Fig. 2c, stacking interactions). We implemented five different stacking regulation variants, utilizing modifications of DNA strands B and D (Fig. 2c, B1 to B5 and D1 to D5). The efficacy of the YES gate was evaluated using a fluorescence assay, where the activation of DNAzyme cleavage by stacking interactions resulted in a notable increase in fluorescence intensity, due to the separation of the fluorophore FAM from the quencher BHQ-1 (Fig. 2b). These results indicate that the optimal signal-to-noise ratio of the YES gate was obtained when using DNA strands B1 and D1 (stacking interaction 1) (Fig. 2d). The relevant regulatory mechanism analysis, structural simulations and dynamic fluorescence results can be found in Supplementary Figs. 2–11 and 22–25.

In the optimal YES gate experiments (Fig. 2e), the fluorescence intensity gradually increased upon introducing trigger D1, indicating that the stacking interaction indeed allosterically activated DNAzyme-1. In contrast, no significant fluorescence increase was generated without trigger D1 (Fig. 2e). The YES gate reaction triggered by stacking interaction 2 was verified by native polyacrylamide gel electrophoresis (PAGE). When introducing trigger D2, the disassembly of DNAzyme-1/substrate A was clearly observed for the stacking interaction-triggered DNAzyme-1 (Fig. 2f, lane 7). Moreover, the YES gate was also tested by the experiments with varying trigger concentrations, where the fluorescence intensity gradually increased as the concentrations of trigger D2 increased from 0 to 200 nM (Fig. 2g). The simulation results can be found in Supplementary Figs. 52–55.

Dual-input SMALL logic function switching

Dual-input SMALL logic function switching was performed by programming two stacking interactions on one DNAzyme to allosterically control the cleavage (Fig. 3a, d). Taking advantage of the flexible DNAzyme construct, we implemented down and up SMALL logic function switching by programming stacking interactions at different structural sites (Fig. 3a, d). In the down SMALL regulation, two logic gates with the functions of AND and OR were constructed by using DNAzyme-2 and DNAzyme-3, respectively. (Fig. 3a).

In the down regulation, the SMALL AND gate was constructed by using DNAzyme-2 (B/C1) to respond to two stacking input D and E1 (Fig. 3a). In the experiments, DNAzyme-2 bound to substrate A to form complex (B/C1/A), a metastable DNAzyme cleavage state (Fig. 3a). Initially, DNAzyme-2 was in the OFF state when stacking interaction 1 was set as site 0 nt at the right side of DNAzyme-2. Introducing either stacking input D or E1 cannot give any significant fluorescent signal, indicating that DNAzyme-2 can not be activated either by a single stacking interaction (Fig. 3b). Only when both stacking inputs were introduced simultaneously, was the fluorescence intensity observed to increase significantly (Fig. 3b). These experimental results demonstrated that DNAzyme-2 can be allosterically activated by both two stacking interactions to cleavage the substrate. The result of the AND gate was also verified by native PAGE (Fig. 3b).

Based on SMALL method, the DNA logic gate function was designed to switch from AND to OR by simply adjusting the stacking interaction from site 0 nt to 2 nt, with only a 2 nt molecular architecture change between DNAzyme-2 and DNAzyme-3 (Fig. 3a). In the results, introducing either stacking input D or E2 can produced a significant fluorescence increase, indicating that just one stacking interaction was able to allosterically trigger DNAzyme cleavage in the OR gate (Fig. 3c). The OR gate was also verified by native PAGE (Fig. 3c). The experimental results well confirm the SMALL based function switching between AND and OR gates, which simply requiring a 2 nt molecular architecture change of the DNAzyme (Fig. 3a). Similarly, in the up SMALL regulation, the function switching of AND and OR gates were implemented by programming stacking interactions 3 and 4 in DNAzyme-4 and DNAzyme-5, respectively (Fig. 3d). The results of the upregulating AND and OR gates are shown in Figs. 3e and f. More information on related structural design, performance analysis, systematic optimization, and simulations can be found in Supplementary Figs. 12–27 and 56–57.

Multi-input SMALL logic function switching

We further proposed multi-input SMALL logic function switching by cooperatively programming multiple stacking interactions to activate the DNAzyme cleavage (Fig. 4a, b). For the multi-input SMALL logic gates, their logic functions could be switched by precisely setting the combination of the four stacking interactions (Fig. 4a, c). For example, to perform a four-input AND gate, the four stacking interactions was set as combination-1 {2, 0, 1, 1 nt} (Fig. 4a, c). Here, DNAzyme-6 was designed as metastable to ensure that the cleavage was activated, only when all four stacking interactions were introduced simultaneously (Fig. 4b and Supplementary Fig. 28). In the experimental results, a significant increase in fluorescence intensity was only observed when all four stacking inputs present simultaneously (Fig. 4d). It should be noted that no significant fluorescence increase was observed when any of the four stacking interactions were missing (Fig. 4d). Similarly, the native PAGE results also confirmed the well performance of the four-input AND functions, where the cleavage was clearly observed only when all four stacking interactions were present simultaneously (Fig. 4e).

Based on the initial four-input SMALL AND gate, the logic functions of the multi-input logic gates were switched by precisely setting the stacking interactions combinations to perform different three-input AND gate-1, -2, and -3 (Fig. 4f–h). For instance, by programming the stacking interactions setting as combination-2 {2, 0, 0, 2 nt}, the logic function was switched to three-input AND gate-1, triggered by stacking inputs (E, K, L) (Fig. 4f). In the experiments of three-input AND gate-1, a significant increase in fluorescence intensity was observed when stacking inputs (E, K, L) were simultaneously introduced to allosterically activate the cleavage. Conversely, no significant increase in fluorescence was detected when any of the three stacking interactions were absent. (Fig. 4f). Similarly, when adjusting as combination-3 {1, 1, 2, 0 nt}, two kinds of the specific three-input stacking AND gates-2 and -3 were performed, triggered by stacking inputs (D, E, K) and (D, E, L) (Fig. 4g, h). In addition, we conducted tests on multiple additional four-input SMALL gates with different stacking interactions combinations. The results demonstrate that the performance of these four-input SMALL gates can also be diversely modulated by setting the stacking interactions combinations (Supplementary Figs. 29–37).

Cascaded SMALL network logic function switching

To investigate whether the SMALL based function switching can be cooperatively integrated in cascaded DNA logic networks, the allosteric stacking interactions are programmed in a hierarchical manner. We firstly constructed a three-level cascaded SMALL network (consisting of the three modules as Level-1, -2 and -3, Fig. 5a) to perform logic computing functions (Fig. 5b). In the Level-1 module, DNAzyme-7 was allosterically triggered by introducing DNA strand X to produce stacking interaction 1 (Fig. 5a). Assisted by the stacking interaction, DNAzyme-7 was activated to cut the target DNA substrate (U/T) at a ribonucleotide modification in the middle of DNA strand U. Then, DNA complex (U*/T) was released and served as the trigger for the next Level-2 module (Fig. 5a). In the whole cascaded process, the stacking interaction signal was hierarchically transmitted from Level-1 to Level-3 modules, thus triggering the activity of DNAzyme-9 and producing a fluorescent increase (Fig. 5a).

**Fig. 5: The cascaded SMALL network function switching.**

In the experiments of the three-level cascaded SMALL network, an increase in fluorescence was observed upon introducing DNA strand X (Fig. 5a, c). Additionally, to investigate the activating effects of the initial triggers in the cascaded SMALL networks, we implemented the experiments by varying the concentrations of trigger X (Fig. 5d). The fluorescence intensities were observed to increase gradually with increasing DNA trigger X concentrations (Fig. 5d). Meanwhile, the effects of the cascading level numbers were also considered in the experiments. Here, a two-level cascaded SMALL network was constructed consisting of two modules Level-2 and -3. An attenuation phenomenon of the fluorescence increases was observed in the three-level network when compared with the two-level networks, possibly due to the increased complexity in system and the incomplete substrate cleavage by the stacking triggered DNAzyme in each layer (Fig. 5c and Supplementary Fig. 51). These data indicated that the cascaded stacking interaction signal transmission was indeed initiated from the upstream to the downstream (more experimental details and results can be found in Supplementary Figs. 38–49 and 58–59).

To demonstrate the possibility to perform function switching in the cascaded SMALL networks, we constructed tree topology cascaded SMALL networks to perform YES/YES-AND and YES/YES-OR logic circuits, respectively (Fig. 5e). The logic function switching was implemented by programming the stacking interaction in the downstream gate with only 2 nt molecular architecture change (Fig. 5e). Specifically, the cascaded network performed YES/YES-AND function when stacking interaction 4 was set as site 0 nt, while the function of cascaded network was switched to YES/YES-OR with stacking interaction 5 set as site 2 nt (Fig. 5e). The corresponding fluorescent results demonstrated the performances of the logic function switching in the cascaded SMALL networks (Fig. 5f, g). It should be noted that during the operations of the cascaded SMALL circuits, a certain degree of signal attenuation and fidelity degradation occurs. We attribute this phenomenon to accumulative signal attenuation during top-down signal transmission, resulting from the incomplete cleavage of substrate DNA strands by DNAzyme. Detailed structural and procedural analyses can be found in Supplementary Fig. 50.

SMALL based logic function switching for endogenous microRNA sensing

The SMALL based function switching was further implemented in real cellular environments, triggered by endogenous microRNAs, which are abundantly expressed in various tumor cells and play significant roles in clinical tumor diagnosis and therapeutics^39,40. We selected microRNAs 21, 17 and 30a as the target input signals for the SMALL-based logic operations. In the dual-input SMALL AND gate, microRNAs 21 and 17 were used as endogenous input signals. Assisted by DNA strands S1 and S2, the two DNA/RNA complexes S1/microRNA 21 and S2/microRNA 17 were able to bind with DNAzyme-10 to form stacking interactions 1 and 2 (Fig. 6a). Then, the activated DNAzyme-10 cleaved substrate A, generating a measurable FAM green fluorescence signal (Fig. 6a). Similarly, the functionality of the SMALL logic gate was modulated by programming the stacking interactions of DNAzyme-10. By setting stacking interactions 1 and 2 as {2, 0 nt} and {1, 1 nt}, respectively, the microRNA logic gates were successfully switched to operate as AND and OR gates. (Fig. 6b and Supplementary Figs. 65-66).

**Fig. 6: SMALL-based microRNA logic function switching in living cells.**

The SMALL based logic function switching was first tested in the breast cancer cell line (MCF-7) to confirm their capability in living cells. Initially, specific chemical RNA inhibitors for microRNAs 21 and 17 were used to selectively control the microRNA activity. The SMALL logic gates were then introduced into the cells using Lipofectamine 3000 (Fig. 6a, b). After that, cellular computing results were analyzed using confocal laser scanning microscopy (CLSM) (Fig. 6c, d). In the AND gate configuration with the stacking combination of {2, 0 nt}, a significant fluorescent signal was only observed when both microRNAs 21 and 17 were present, and inhibition of any one microRNA resulted in negligible fluorescence (Fig. 6c). These findings were corroborated by flow cytometry results (Supplementary Fig. 65). Additionally, the OR gate was effectively switched using the stacking combination of {1, 1 nt} to respond to microRNAs 21 and 17 (Fig. 6b-d and Supplementary Fig. 66), demonstrating the simple and flexible function switching of the SMALL platform in living cells.

To further extend the versatility of the SMALL based logic operations, a four-input AND gate was constructed, triggered by endogenous microRNAs (microRNA 21, 17, and 30a) and an exogenous DNA input (DNA strand D) (Fig. 6e, f). Assisted by DNA strands S1, S2, S3 and D, the three DNA/RNA complexes (S1/microRNA 21, S2/microRNA 17 and S3/microRNA 30a) and DNA D engaged with DNAzyme-11, forming stacking interactions 1, 2, 3 and 4, respectively. This activation led to substrate cleavage and the production of a fluorescent signal (Fig. 6e, f). Implemented in a cervical cancer cell line (HeLa), the four-input SMALL AND gate exhibited substantial fluorescence only when all inputs were concurrently present (Fig. 6g, h). In the presence of DNA strand D, introduction of inhibitors for any of the microRNAs significantly reduced fluorescence, confirming the gate’s specificity and sensitivity (Fig. 6g, h). These multi-input logic operations were further validated by additional flow cytometry experiments (Fig. 6i). Details regarding the structural design, performance testing, and anti-interference analysis of the aforementioned miRNA-sensing logic gates are provided in Supplementary Figs. 60–71.

SMALL based logic function switching for gene regulations

Subsequently, a SMALL-based logic function switching system was developed to enable precise gene regulation in cellular environments. To achieve this, we employed antisense oligonucleotides (ASOs) in our system, which are short nucleic acid sequences capable of suppressing gene expression and are commonly utilized in molecular therapeutics for gene regulation^41,42. Specifically, ASOs targeting the GFP protein were integrated into our experiments to be regulated via SMALL-mediated cleavage. This integration established the SMALL-ASO logic networks, which controlled DNAzyme activity and thereby indirectly modulated GFP expression levels (Fig. 7a). In the basic YES logic operation, a one-level SMALL-ASO stacking network was established with the stacking interactions combination set to {2, 2 nt} (Fig. 7a). This network employs endogenous microRNA 21 as an input, which forms the DNA/RNA complex S1/microRNA 21 through hybridization with DNA strand S1. This complex subsequently bound to DNAzyme-12 to establish stacking interaction 1 (Fig. 7a). Then, the stacking activated DNAzyme-12 cleaves the DNA protector A*, releasing D-GFP and thus inhibiting GFP protein expression (Fig. 7a). We initially validated the functionality of the YES gate through in vitro experiments (Supplementary Figs. 72 and 73). Subsequently, by setting the stacking interactions combinations to {2, 0 nt} and {1, 1 nt}, we successfully constructed the AND and OR gates, respectively (Supplementary Figs. 74–76). The functionality of these logic gates was evaluated through both in vitro and cellular experiments, demonstrating their capability to efficiently regulate GFP expression levels.

**Fig. 7: SMALL-ASO logic network function switching to perform programmable gene regulation in living cells.**

Following this, two-level cascaded SMALL-ASO stacking networks were developed to implement logic function switching, enabling more complex regulations of GFP expression. Initially, the function switching mechanism was exclusively established in the upstream gate, which is regulated by microRNAs 21 and 17. Concurrently, the downstream gate was activated by microRNA 30a through an AND logic operation. (Fig. 7b). Initially, the SMALL mechanism was integrated into the upstream gate with two distinct stacking combinations (Fig. 7b). This configuration allows for three types of SMALL-ASO logic operations: YES-AND, OR-AND, and AND-AND, set by the respective stacking combinations of up- {2, 2 nt}, up- {1, 1 nt}, and up- {2, 0 nt}. Similarly, before the operation of the cascaded SMALL-ASO stacking networks, specific chemical RNA inhibitors (targeting microRNAs 21, 17, and 30a) were employed to selectively modulate the activity of these microRNAs. The experimental results confirm that these function switching was effectively implemented in HeLa cells, with molecular architecture change ratios of 0.3% (Fig. 7c). For instance, switching the logic functions from AND-AND to OR-AND involved a mere 2 nt molecular architecture change, accounting for a 0.3% change within the whole 543 nt SMALL-ASO network structure (Fig. 7b, c). Moreover, these results demonstrate the sensitivity and precision of the SMALL-based cellular function switching. Even under identical endogenous conditions of microRNAs 17 and 30a, significantly different computational outputs were achieved (Fig. 7c, TA-2 and TB-2). Additional experimental findings are detailed in Fig. 7c and Supplementary Fig. 79.

The functionality of the SMALL-ASO networks was additionally regulated across both upstream and downstream gates to achieve dual-logic function switching for coordinated gene regulations (Fig. 7d). The SMALL mechanism was implemented both in up- and downstream gates by setting four distinct stacking combinations to implement the function switching (Fig. 7d). This coordinated regulation allowed for the execution of 9 distinct logic functions, managing GFP gene expression in response to 8 unique cellular microRNA conditions (Fig. 7d, e). In total, 72 different SMALL-based gene regulation patterns were achieved in HeLa cells, with molecular architecture change ratios ranging from 0.2% to 0.7% (Fig. 7e). For example, cooperative dual-logic function switching in both upstream and downstream gates facilitated the switch from the AND-AND to the OR-OR configuration. By adjusting the stacking combinations from {2, 0, 2, 0 nt} to {1, 1, 0, 0 nt}, which involved a molecular architecture change ratio of 0.7%, distinct computational outputs were observed, differentiating the AND-AND from the OR-OR SMALL-ASO network (Fig. 7e, TC-1 and TC-9). Relevant cellular experiment data analysis, network structure design, and in vitro performance testing are provided in Supplementary Figs. 77–82.

Discussion

We reported the SMALL strategy, an approach for switching DNA computing functions with minimal molecular architecture changes. Since our work presents a base stacking-mediated allosteric regulation approach, it provides a simplified alternative tool to enrich DNA regulations, which are often manipulated by conventional toehold-mediated DNA strand displacement reactions^31,32. Our design is a simple and universal platform for implementing the DNA computing function switching with several features. (1) The orthogonal design of the allosteric DNAzyme gate regulation. (2) Simple switch of DNA logic gate function. (3) Cooperative stacking interactions with high programmability. (4) Well biological applicability.

Furthermore, related potential applications may be worth consideration in two aspects: (1) Programmable molecular circuits for genome detection. The SMALL strategy may be used to detect genome, for example, DNA breaks (Double-Strand Breaks, DSBs) by constructing base stacking interaction sites that are similar to the structure of DSBs. Therefore, SMALL could enable the simple design of programmable DNA circuits for sensitive sensing of DSBs. (2) Bridging enzyme networks with DNA-based computing. Modern enzyme engineering has produced diverse, programmable enzymes (e.g., restriction enzymes, CRISPR-Cas systems) capable of generating targeted cleavage sites in DNA structures. The resulting structures from enzymatic cleavage are very similar to the base stacking interaction sites in SMALL circuits. Therefore, SMALL circuit may provide a framework to transform enzyme-generated cleavage signals into DNA computing processes. This integration would establish a critical bridge between natural enzymatic systems and synthetic molecular information processing.

We believe that the SMALL mechanism offers an alternative and simple tool to implement DNA computing function switching. Leveraging the simplicity and programmability inherent to the SMALL strategy, our method holds promise for constructing more programmable stacking-mediated DNA networks capable of executing complex function switching. Therefore, we anticipate that this strategy will significantly enrich the regulation dimensions of nucleic acid systems, and open up new routes for developing new technologies for specific molecular sensing and gene engineering applications.

Methods

Materials

All oligonucleotides, including synthetic microRNAs, were obtained from Sangon Biotech (China). The unmodified DNA strands underwent purification through polyacrylamide gel electrophoresis (PAGE), while the DNA strands labeled with fluorophores, quenchers, and ribonucleotides were purified using high-performance liquid chromatography (HPLC). All DNA strand sequences are listed in Supplementary tables 1-10. The DNA sequences utilized in this study were designed and optimized with the help of NUPACK software (http://www.nupack.rog), ensuring minimal crosstalk between unwanted domains⁴³. Magnesium acetate tetrahydrate (Mg (CH₃COO)₂·4H₂O) was sourced from Sinopharm Chemical Reagent (Catalogue No. 30110518, China), while 50 × TAE buffer was procured from Beijing Solarbio Science & Technology (Catalogue No. T1060, China). The experiments utilized ultrapure water produced by Beijing Feimo Bio-Technology (Catalogue No. FB19237, China). To prepare for experimental procedures, all DNA strands were kept in ultrapure water at a concentration of around 100 μM and stored at –20 °C prior to their application. The concentrations of DNA and microRNA were measured using a Nanodrop-2000 (Thermo Fisher Scientific, Catalogue No. ND-2000C, USA) at a wavelength of 260 nm. For all experiments conducted without cells, the buffer used was 1 × TAE containing 11 mM Mg (CH₃COO)₂.

Fetal bovine serum (FBS, Catalogue No. 018-150), Dulbecco’s modified Eagle medium (DMEM, Catalogue No. 319-005), and phosphate-buffered saline (PBS, Catalogue No. 311-010) were sourced from Wisent Corporation (Canada). The human breast cancer cells (MCF-7 and MCF-7-GFP) and human cervical cancer cells (HeLa and HeLa-GFP) were obtained from the Cell Bank of the Shanghai Institute of Biochemistry and Cell Biology. All microRNA inhibitors (Catalogue No. miR20000070-1-5, miR20000076-1-5, miR20000087-1-5) were obtained from RiboBio (China).

Assembly procedure of DNA complexes

The specific equimolar concentrations (e.g., 2 μM) single-stranded DNA (ssDNA) strands were combined in 1 × TAE/Mg²⁺ buffer to generate the DNA complexes, which were subsequently annealed with a PCR Thermal Cycler (Bio-Rad, Catalogue No. 1861096, USA). The PCR procedure consisted of raising the temperature of the samples to 95 °C for a duration of 5 min, followed by a gradual cooling to 25 °C at a steady rate of 5 °C every 5 min. Following the annealing process, all DNA complexes that underwent purification steps were kept at room temperature until they were utilized. The ssDNA samples (e.g., 2 μM) were employed without any further processing.

Fluorescence assay

All kinetic fluorescence experiments were performed in 1 × TAE/Mg²⁺ buffer utilizing real-time fluorescence PCR (Agilent Technologies, Catalogue No. G8830A, USA). The FAM optical cartridge had an excitation wavelength of 462 nm and an emission wavelength of 526 nm. In a standard kinetic experiment, a 30 μL sample was used for the test procedure, with measurements taken every 3 min. The kinetic measurements commenced immediately after the trigger was added. Fluorescence measurements for the background were conducted by using buffer in place of the trigger. The fluorescence experiments were performed three times, and the average results were presented. All experiments were performed using a one-pot approach.

PAGE experiments

The validation of the SMALL reactions was conducted using native PAGE. In these experiments, a 12 % gel was prepared with 1 × TAE/Mg²⁺ buffer containing 11 mM Mg (CH₃COO)₂. At the beginning, all samples were combined with glycerin to achieve a final concentration of 6% (v/v), then placed into the gel and subjected to electrophoresis at 4 °C for 1.5 h at a voltage of 100 V. The gels were subsequently treated with Stain-All solution for 2 h and visualized with a Canon LIDE 100 scanner (Canon, Catalogue No. 9139B010, Japan). The DNA complex labeled with FAM fluorophores was visualized using a Florchem FC2 gel scanner (Alpha Innotech, Catalogue No. 92-1620, USA) under ultraviolet light, without the application of any staining. All SMALL reactions verified by PAGE were performed as pooled samples in 1 × TAE/Mg²⁺ buffer over 12 hours. The DNA strand concentrations and reaction temperatures are detailed in the corresponding figure legends.

Calculating fluorescence

Original fluorescence data were standardized using the standardized data processing method to enable comparison and analysis of the fluorescence results. The fluorescence results were standardized using the equation:

$${{\rm{Normalized}}}\; {{\rm{fluorecence}}}\; {{\rm{change}}}\; {{\rm{ratio}}}=\frac{{F}_{R}-{F}_{i}}{{F}_{i}}$$

(1)

Here, ${F}_{i}$ denotes the fluorescence intensity in the experiments without being triggered. ${F}_{R}$ is the fluorescence intensity of the experiments triggered by input. The values of ${F}_{R}$ in the reactions are always larger than ${F}_{i}$. Thus, all fluorescence results range between 0 and 1.

Simulations of the reaction kinetics

The dynamic mechanism of the stacking interaction was mathematically represented using state-space models that use state variables to describe the interaction system by a set of first-order differential equations⁴⁴. Initially, all reaction formulas were mathematically modeled based on the state-space forms in a uniform schema. Next, nonlinear grey-box identification was used to estimate the unknown coefficients in the above mathematical model to fit given experimental response trajectories⁴⁵. Finally, the entire reaction process was simulated using the updated coefficients. More details of the simulation and calculations can be found in Supplementary Figs. 52–59.

Kinetic fluorescence measurements of the SMALL systems

The basic SMALL regulation module experiments were performed in 1 × TAE/Mg²⁺ buffer containing DNAzyme-1 (B/C) (200 nM) and substrate A (200 nM), and incubated at 25 °C for 5 min using the real-time fluorescence PCR. Kinetic fluorescence measurements were initiated by adding trigger D (200 nM). Fluorescence changes were then recorded by the real-time fluorescence PCR for 150 min at 25 °C. For the basic module experiments with other stacking interaction sites, the DNA concentrations, reaction conditions, and operational procedures were identical to those described above.

The downregulation SMALL AND gate reaction was performed in the reaction buffer containing DNAzyme-2 (B/C1) (200 nM) and substrate A (200 nM), and incubated at 25 °C for 5 min. Triggers D (200 nM) and E1 (200 nM) were added to the gate sample to induce kinetic fluorescence measurements. For the downregulation SMALL OR gate reaction, the reaction buffer containing DNAzyme−3 (B/C2) (200 nM) and substrate A (200 nM) was mixed at 25 °C for 5 min. Kinetic fluorescence measurements were initiated by adding trigger D (200 nM) or trigger E2 (200 nM). Similarly, the upregulation SMALL AND gate reaction was conducted in the reaction buffer containing DNAzyme-4 (F/G/H1) (200 nM) and substrate A (800 nM). The fluorescence measurements were initiated by adding triggers K (200 nM) and L1 (200 nM). In the upregulation SMALL OR gate reaction, the reaction buffer containing DNAzyme-5 (F/G/H2) (200 nM) was mixed with substrate A (800 nM) for 5 min at 25 °C. The kinetic fluorescence measurements were initiated by adding trigger K (200 nM) or L2 (200 nM). In these logical reactions, the kinetic fluorescence measurements were recorded by the real-time fluorescence PCR for 150 min at 25 °C. For the dual-input model experiments with other stacking interaction sites, the DNA concentrations, reaction conditions, and operational procedures were identical to those described above.

In the multi-input SMALL logic gates, DNAzyme (200 nM) was mixed with substrate A (800 nM) in 1 × TAE/Mg²⁺ buffer and incubated at 30 °C for 5 min in the real-time fluorescence PCR instrument. Triggers D (200 nM), E (200 nM), K (200 nM) and L (200 nM) were added to the corresponding modules to initiate the kinetic fluorescence measurements. Kinetic fluorescence measurements were recorded by the real-time fluorescence PCR for 900 min at 30 °C.

In the two-level SMALL networks, DNAzyme-8 (R/P/Q) (200 nM) was mixed with DNA complex (N/M) (400 nM/200 nM) to form the level-2 module, and DNAzyme-9 (F/G/H) (200 nM) was mixed with substrate A (800 nM) to form the level-3 module. Trigger T (200 nM) was added into the network, and kinetic fluorescence measurements were recorded with the real-time fluorescence PCR for 900 min at 30 °C. In the three-level SMALL networks, the level-2 and level−3 modules were similar to those used in the two-level networks, whereas DNAzyme-7 (W/Y/Z) (200 nM) was mixed with complex (U/T) (200 nM/ 200 nM) to form the level-1 module. Kinetic fluorescence measurements were initiated with the addition of trigger X (200 nM), and fluorescence data were recorded by the real-time fluorescence PCR for 900 min at 25 °C.

In the YES/YES-AND and YES/YES-OR SMALL networks, DNAzyme (R/P/Q) (100 nM) was mixed with DNA complex (N/M) (200 nM/100 nM) to form the first YES gate, and DNAzyme (W/Y/Z) (100 nM) was mixed with DNA complex (J/I) (200 nM/100 nM) to form the second YES gate. DNAzyme (F/G/H) (100 nM) was mixed with substrate A (400 nM) to form the AND or the OR gate. Triggers X (100 nM) and T (100 nM) were added into the networks, and kinetic fluorescence measurements were recorded with the real-time fluorescence PCR for 900 min at 25 °C.

In the SMALL-ASO and SMALL sensing gates, DNAzyme complex (200 nM) was mixed with DNA substrate (800 nM) in 1 × TAE/Mg²⁺ buffer and incubated at 25 °C for 5 min in the real-time fluorescence PCR instrument. Then, microRNA-21 (200 nM), microRNA-17 (200 nM), and microRNA-30a (200 nM) were added to the corresponding gates to initiate the kinetic fluorescence measurements. Fluorescence changes were then recorded by the real-time fluorescence PCR for 150 min at 25 °C.

Cell culture and transfection of DNA systems

MCF-7, MCF-7 GFP, HeLa and HeLa GFP cells were cultured in Dulbecco’s Modified Eagle Medium (DMEM) medium supplemented with 10% fetal bovine serum (FBS) and 1% penicillin-streptomycin. The cells were maintained in a 5% CO₂ atmosphere at 37 °C. After seeding the cells into culture dishes for different times, various DNA systems (with a final DNAzyme concentration of 200 nM and substrate concentration of 800 nM) were introduced into the cells via the Lipofectamine 3000 kit (Thermo Fisher Scientific, Catalogue No. L3000015, USA), following the manufacturer’s instructions. To modulate miRNA levels within the living cells, overdose miRNA inhibitors were transfected via Lipofectamine 3000 kit prior to use.

Confocal microscopy analysis

MCF-7, MCF-7 GFP, HeLa and HeLa GFP tumor cells were seeded in 35 mm confocal microscopy dishes (Cellvis, Catalogue No. D35-20-1-N, USA) and cultured to approximately 50% confluency over 24 h before treatment. The cells were then transfected with various DNA systems via the Lipofectamine 3000 kit. Then the DNA systems were co-incubated with different cells (with a final DNAzyme concentration of 200 nM and substrate concentration of 800 nM), and the cell incubation time specifically changed in different experiments. In the fluorescence analysis experiment, the cell incubation time was 12 h. In the GFP-gene regulation experiment, the cell incubation time was about 48 h. Imaging was performed using an Olympus FV3000 laser scanning confocal microscope (Olympus Corporation, Catalogue No. FV30SIL, USA). The captured images were subjected to further analysis utilizing the Imaris microscopy image analysis software.

Flow cytometry analysis

MCF-7, MCF-7 GFP, HeLa and HeLa GFP cells were seeded in 12-well plates and cultured with the same conditions as described above to ~50% confluence over 12 h before treatment. Then the DNA systems (with a final DNAzyme concentration of 200 nM and substrate concentration of 800 nM) were transfected using the Lipofectamine 3000 kit and co-incubated with different cells, with incubation times varying depending on the experiment. For fluorescence analysis experiment, the cell incubation time was 12 h while for the GFP-gene regulation experiment, it was ~48 h. After incubation, the cells were trypsinized for 2 min to obtain suspended cells, which were then washed three times with PBS. Finally, the FAM green fluorescence and GFP signal were detected using a CytoFLEX system (Beckman Coulter, Catalogue No. C30720, USA) with an excitation wavelength of 488 nm and at least 10,000 cells were recorded for each measurement. A detailed analysis of the flow cytometry gating strategies is provided in Supplementary Fig. 7.4 in the Supplementary Information.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.

Data availability

All data in the main text and Supplementary Information were obtained from three independent experiments, and represent the average results of these experiments. Source data are provided with this paper.

References

Cahill, M. K. et al. Network-level encoding of local neurotransmitters in cortical astrocytes. Nature 629, 146–153 (2024).
Article ADS CAS PubMed PubMed Central Google Scholar
Li, H. S. et al. Multidimensional control of therapeutic human cell function with synthetic gene circuits. Science 378, 1227–1234 (2022).
Article ADS CAS PubMed PubMed Central Google Scholar
Yen, H. Y. et al. Mass spectrometry captures biased signalling and allosteric modulation of a G-protein-coupled receptor. Nat. Chem. 14, 1375–1382 (2022).
Article CAS PubMed PubMed Central Google Scholar
Mathy, C. J. P. et al. A complete allosteric map of a GTPase switch in its native cellular network. Cell Syst. 14, 237–246.e237 (2023).
Article CAS PubMed PubMed Central Google Scholar
Ying, W. et al. Structure and function of the Arabidopsis ABC transporter ABCB19 in brassinosteroid export. Science 383, eadj4591 (2024).
Article CAS PubMed Google Scholar
Williams, J. Z. et al. Precise T cell recognition programs designed by transcriptionally linking multiple receptors. Science 370, 1099–1104 (2020).
Article ADS CAS PubMed PubMed Central Google Scholar
Niu, X. et al. Pseudoknot length modulates the folding, conformational dynamics, and robustness of Xrn1 resistance of flaviviral xrRNAs. Nat. Commun. 12, 6417 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Meijer, L. H. H. et al. Hierarchical control of enzymatic actuators using DNA-based switchable memories. Nat. Commun. 8, 1117 (2017).
Article ADS PubMed PubMed Central Google Scholar
Schaffter, S. W. et al. Standardized excitable elements for scalable engineering of far-from-equilibrium chemical networks. Nat. Chem. 14, 1224–1232 (2022).
Article CAS PubMed Google Scholar
Zheng, L. L. et al. Enthalpy and entropy synergistic regulation-based programmable DNA motifs for biosensing and information encryption. Sci. Adv. 9, eadf5868 (2023).
Article CAS PubMed PubMed Central Google Scholar
Tang, Q. et al. Multi-mode reconfigurable DNA-based chemical reaction circuits for soft matter computing and control. Angew. Chem. Int. Ed. Engl. 60, 15013–15019 (2021).
Article CAS PubMed Google Scholar
Xiong, X. et al. Optochemical control of DNA-switching circuits for logic and probabilistic computation. Angew. Chem. Int. Ed. Engl. 60, 3397–3401 (2021).
Article CAS PubMed Google Scholar
Kawamata, I. et al. Autonomous assembly and disassembly of gliding molecular robots regulated by a DNA-based molecular controller. Sci. Adv. 10, eadn4490 (2024).
Article CAS PubMed PubMed Central Google Scholar
Fu, R., Hou, J., Wang, Z. & Xianyu, Y. DNA molecular computation using the CRISPR-mediated reaction and surface growth of gold nanoparticles. ACS Nano 18, 14754–14763 (2024).
Article CAS PubMed Google Scholar
Liu, Y. et al. Erasable and field programmable DNA circuits based on configurable logic blocks. Adv. Sci. (Weinh.) 11, e2400011 (2024).
PubMed Google Scholar
Arter, W. E. et al. Digital sensing and molecular computation by an enzyme-free DNA circuit. ACS Nano 14, 5763–5771 (2020).
Article CAS PubMed Google Scholar
Sarraf, N., Rodriguez, K. R. & Qian, L. Modular reconfiguration of DNA origami assemblies using tile displacement. Sci. Robot. 8, eadf1511 (2023).
Article PubMed Google Scholar
Douglas, S. M., Bachelet, I. & Church, G. M. A logic-gated nanorobot for targeted transport of molecular payloads. Science 335, 831–834 (2012).
Article ADS CAS PubMed Google Scholar
Ji, W. et al. Encoding signal propagation on topology-programmed DNA origami. Nat. Chem. 16, 1408–1417 (2024).
Article CAS PubMed Google Scholar
Chatterjee, G., Dalchau, N., Muscat, R. A., Phillips, A. & Seelig, G. A spatially localized architecture for fast and modular DNA computing. Nat. Nanotechnol. 12, 920–927 (2017).
Article ADS CAS PubMed Google Scholar
Wang, C., O’Hagan, M. P., Neumann, E., Nechushtai, R. & Willner, I. Integration of photocatalytic and dark-operating catalytic biomimetic transformations through DNA-based constitutional dynamic networks. Nat. Commun. 12, 4224 (2021).
Article ADS CAS PubMed PubMed Central Google Scholar
Liang, Y. et al. Programming conformational cooperativity to regulate allosteric protein-oligonucleotide signal transduction. Nat. Commun. 14, 4898 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Huang, F. et al. DNA branch migration reactions through photocontrollable toehold formation. J. Am. Chem. Soc. 135, 7967–7973 (2013).
Article ADS CAS PubMed PubMed Central Google Scholar
Yang, S. et al. Light-activated signaling in DNA-encoded sender-receiver architectures. ACS Nano 14, 15992–16002 (2020).
Article CAS PubMed PubMed Central Google Scholar
Hartmann, D., Chowdhry, R., Smith, J. M. & Booth, M. J. Orthogonal light-activated DNA for patterned biocomputing within synthetic cells. J. Am. Chem. Soc. 145, 9471–9480 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Haydell, M. W., Centola, M., Adam, V., Valero, J. & Famulok, M. Temporal and reversible control of a DNAzyme by orthogonal photoswitching. J. Am. Chem. Soc. 140, 16868–16872 (2018).
Article ADS CAS PubMed Google Scholar
Guo, Y. et al. pH-controlled detachable DNA circuitry and its application in resettable self-assembly of spherical nucleic acids. ACS Nano 14, 8317–8327 (2020).
Article CAS PubMed Google Scholar
Amodio, A. et al. Rational design of pH-controlled DNA strand displacement. J. Am. Chem. Soc. 136, 16469–16472 (2014).
Article ADS CAS PubMed Google Scholar
Sun, X., Yao, D. & Liang, H. pH-controlled resettable modular DNA strand-displacement circuits. Nano Lett. 23, 11540–11547 (2023).
Article ADS CAS PubMed Google Scholar
Porchetta, A., Idili, A., Vallée-Bélisle, A. & Ricci, F. General strategy to introduce pH-induced allostery in DNA-based receptors to achieve controlled release of ligands. Nano Lett. 15, 4467–4471 (2015).
Article ADS CAS PubMed PubMed Central Google Scholar
Jung, J. K., Archuleta, C.M., Alam, K.K. & Lucks, J.B. Programming cell-free biosensors with DNA strand displacement circuits. Nat. Chem. Biol. 18, 385–393 (2022).
Article CAS PubMed PubMed Central Google Scholar
Wang, F. et al. Implementing digital computing with DNA-based switching circuits. Nat. Commun. 11, 121 (2020).
Article ADS PubMed PubMed Central Google Scholar
Qi, M. et al. Meta-DNA strand displacement for sub-micron-scale autonomous reconfiguration. J. Am. Chem. Soc. 145, 16812–16820 (2023).
Article ADS CAS PubMed Google Scholar
Nikitin, M. P. Non-complementary strand commutation as a fundamental alternative for information processing by DNA and gene regulation. Nat. Chem. 15, 70–82 (2023).
Article CAS PubMed Google Scholar
Liao, Y. et al. A versatile and convenient tool for regulation of DNA strand displacement and post-modification on pre-fabricated DNA nanodevices. Nucleic Acids Res 51, 29–40 (2023).
Article CAS PubMed Google Scholar
Su, H., Xu, J., Wang, Q., Wang, F. & Zhou, X. High-efficiency and integrable DNA arithmetic and logic system based on strand displacement synthesis. Nat. Commun. 10, 5390 (2019).
Article ADS PubMed PubMed Central Google Scholar
Song, T. et al. Fast and compact DNA logic circuits based on single-stranded gates using strand-displacing polymerase. Nat. Nanotechnol. 14, 1075–1081 (2019).
Article ADS CAS PubMed Google Scholar
Lv, H. et al. DNA-based programmable gate arrays for general-purpose DNA computing. Nature 622, 292–300 (2023).
Article ADS CAS PubMed Google Scholar
Lin, N. et al. Spatially localized entropy-driven evolution of nucleic acid-based constitutional dynamic networks for intracellular imaging and spatiotemporal programmable gene therapy. J. Am. Chem. Soc. 146, 20685–20699 (2024).
Article ADS CAS PubMed PubMed Central Google Scholar
Li, Z. et al. Colocalization of protein and microRNA markers reveals unique extracellular vesicle subpopulations for early cancer detection. Sci. Adv. 10, eadh8689 (2024).
Article CAS PubMed PubMed Central Google Scholar
Su, X. et al. Efficient inhibition of SARS-CoV-2 using chimeric antisense oligonucleotides through RNase L activation*. Angew. Chem. Int. Ed. Engl. 60, 21662–21667 (2021).
Article CAS PubMed PubMed Central Google Scholar
Terada, C. et al. Dynamic and static control of the off-target interactions of antisense oligonucleotides using toehold chemistry. Nat. Commun. 14, 7972 (2023).
Article ADS CAS PubMed PubMed Central Google Scholar
Zadeh, J. N. et al. NUPACK: analysis and design of nucleic acid systems. J. Comput. Chem. 32, 170–173 (2011).
Article ADS CAS PubMed Google Scholar
Nise, N. S. Control Systems Engineering (Wiley, Hoboken, NJ, 2014).
Bohlin, T. P. Practical Grey-box Process Identification: Theory and Applications (Springer, London, 2006).

Download references

Acknowledgements

This work was supported by the National Key Research and Development Program of China (2024YFF1206200 to C. Z., 2023YFF1205600 to J. Y., 2022YFC2403203-3, 2023YFC3403100, 2022YFA1207300 to S. P. L.), the National Natural Science Foundation of China (T2525003 and 62273008 to C. Z., 62472171 to J. Y.), Pre-Research Project (31511090301 to C. Z.), the Fund of Kunpeng & Ascend Center of Excellence (Peking University) to C. Z.

Author information

These authors contributed equally: Yongpeng Zhang, Bozhao Li.

Authors and Affiliations

Yanzhao Electric Power Laboratory, School of Control and Computer Engineering, North China Electric Power University, Beijing, China
Yongpeng Zhang, Xuan Liu, Shi Liu & Jing Yang
School of Computer Science, Peking University, Beijing, China
Yongpeng Zhang, Xuan Liu & Cheng Zhang
CAS Key Laboratory for Biomedical Effects of Nanomaterials and Nanosafety, CAS Center of Excellence in Nanoscience, National Center for Nanoscience and Technology, Beijing, China
Bozhao Li, Guangjun Nie & Suping Li
Institutes of Biomedical Sciences, Inner Mongolia University, Hohhot, China
Bozhao Li
College of Computer Science, Shenyang Aerospace University, Shenyang, China
Xuedong Zheng
Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and Emory University, Emory University School of Medicine, Atlanta, GA, USA
Yonggang Ke
Department of Chemistry, Emory University, Atlanta, GA, USA
Yonggang Ke

Authors

Yongpeng Zhang
View author publications
Search author on:PubMed Google Scholar
Bozhao Li
View author publications
Search author on:PubMed Google Scholar
Xuan Liu
View author publications
Search author on:PubMed Google Scholar
Xuedong Zheng
View author publications
Search author on:PubMed Google Scholar
Shi Liu
View author publications
Search author on:PubMed Google Scholar
Guangjun Nie
View author publications
Search author on:PubMed Google Scholar
Jing Yang
View author publications
Search author on:PubMed Google Scholar
Yonggang Ke
View author publications
Search author on:PubMed Google Scholar
Suping Li
View author publications
Search author on:PubMed Google Scholar
Cheng Zhang
View author publications
Search author on:PubMed Google Scholar

Contributions

C.Z. and Y.Z. conceived the research. Y.Z. and B.L. performed the majority of the experiments and analyzed data. C.Z., Y.Z., S.Li, J.Y., B.L., X.L., X.Z. and S.Liu participated in data analysis and discussions. C.Z., S.Li, J.Y., G.N. and Y.K. provided supervision. C.Z., S.Li, J.Y. and Y.K. wrote the original manuscript. C.Z., S.Li, J.Y., Y.K. and G.N. reviewed and edited the manuscript.

Corresponding authors

Correspondence to Jing Yang, Yonggang Ke, Suping Li or Cheng Zhang.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Communications thanks the anonymous reviewer(s) for their contribution to the peer review of this work. A peer review file is available.

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary information

Supplementary Information (download PDF )

Reporting Summary (download PDF )

Transparent Peer Review file (download PDF )

Source data

Source Data (download XLSX )

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Cite this article

Zhang, Y., Li, B., Liu, X. et al. DNA computing function switching by programming base stacking interactions with minimal molecular architecture changes. Nat Commun 16, 9369 (2025). https://doi.org/10.1038/s41467-025-64406-x

Download citation

Received: 12 March 2025
Accepted: 12 September 2025
Published: 23 October 2025
Version of record: 23 October 2025
DOI: https://doi.org/10.1038/s41467-025-64406-x