A pH-regulated camouflage strategy for access control in molecular systems

Abstract

Continuous advancements in DNA nanotechnology have resulted in innovative information security strategies utilizing biomolecular-based encryption mechanisms to achieve effective protection and reliable storage. However, existing molecular encryption schemes are limited in their practical expansion due to their focus on the precise design and strict management of the key itself while neglecting the impact of complex environments, causing them to suffer from insufficient concealment and poor adaptability to complex access control scenarios. Here, we propose a pH-regulated camouflage strategy that significantly enhances the encryption concealment of molecular systems. Using pH-regulated logical operation transformations, we establish a switchable cascaded logic operation strategy that implements a two-layer DNA molecular firewall access control system. By integrating two distinct camouflage rules, the system constructs four differentiated access modes. This design effectively prevents unauthorized access, conceals true input addresses, and deceives potential attackers, thereby ensuring internal information security. This approach provides a simple method for the design and management of molecular systems and is expected to be applied in fields such as information security and environmental monitoring.

---

New concepts

This work demonstrates a novel concept in DNA molecular cryptography: using DNA triplex strands to construct molecular access control systems for encryption. Unlike previous methods that relied on precise key design and strict management for molecular encryption, this approach utilizes DNA triplex strands that are responsive to complex environmental changes. The triplex strand switches act as firewalls for the molecular system, introducing competitive suppression toeholds to masquerade the true access address. This not only defends against external attacks but also allows the system to output differentiated information, confusing attackers. This work overcomes the problems of insufficient encryption concealment and difficulty in adapting to access control requirements in dynamic scenarios, providing new insights into building secure, adaptive molecular systems. Furthermore, its excellent pH response and masquerading capabilities may expand its potential applications in environmental detection.

1. Introduction

In the current era of information explosion and technological iteration, security protection has extended from physical spaces to the molecular level. Modern cryptography, a cornerstone of information security, is primarily grounded in the principles of computational complexity theory.¹ As computational power and quantum computing technologies continue to advance, the scale of key management in modern cryptography is also expanding exponentially; thus, related security issues face significant challenges.² Given these challenges, leveraging the inherent complexity of molecular systems to construct new cryptographic paradigms has become a promising approach for improving cryptographic security. Therefore, exploring cryptographic strategies that adapt to the properties of biomolecules and complex biochemical reactions is a reliable strategy for enhancing the risk resistance of cryptographic systems. As the natural carriers of genetic information, DNA molecules possess diverse properties, including high specificity,^3,4 addressability,^5,6 and programmability.^7,8 These properties enable their broad application in fields such as information encryption,^9,10 nanodevices,^11,12 logical computation,^13,14 and biosensing.^15,16

The four nucleobases utilized in DNA can form billions of sequence combinations through the Watson–Crick base pairing principle,^17,18 creating an almost infinite key space for security protection. Adapting this feature as a security strategy has been widely explored by researchers. For example, programmable encryption strategies that utilize long-chain DNA synthesis and sequence encoding enable the decoding and retrieval of encrypted text and image data based on user permissions.¹⁹ Biomolecule-driven two-factor authentication strategies employ specific DNA strands and enzymes to control access to secure devices.²⁰ Furthermore, time-controllable molecular authentication strategies ensure that system authorization and access are time-limited, thereby achieving spontaneous deactivation.²¹ However, these schemes focus on precise design and strict management of the keys while ignoring the impact of complex environments on encryption systems. This oversight leads to insufficient encryption concealment and makes it difficult to adapt access control in dynamic scenarios, limiting the practical expansion of molecular encryption. Therefore, there is an urgent need to develop an access control method that is adaptable to complex environments, further enhancing the concealment and complexity of encryption techniques, to expand the applications of molecular encryption technologies based on traditional key-dependent schemes and strengthen the security and protective capabilities of molecular systems. To address the critical need for environmental adaptability in encryption systems, we have directed our attention to pH-responsive DNA triplexes. These structures can convert external environmental signals directly into programmable molecular responses,^22,23 providing a molecular basis for constructing environmentally driven encryption systems. The DNA triplex is a unique structure formed via Hoogsteen interactions,²⁴ composed of oligodeoxynucleotides linked through pyrimidine-rich triplex-forming sequences.^25–27 In acidic environments, triplexes can form C–G–C⁺ hydrogen bonds via the protonation of cytosine, giving it unique pH responsiveness,^28–30 which has garnered marked attention from researchers. Examples of this approach include detachable DNA circuits,³¹ photochemically controlled DNA switch circuits,^32,33 and pH-controllable polymorphic DNA switch circuits.³⁴ Therefore, DNA triplexes can utilize their intrinsic pH responsiveness to adjust their structural conformation dynamically under varying pH conditions. This capability effectively addresses the issues of insufficient encryption stealth and poor adaptability to complex access control scenarios. Consequently, leveraging triplexes to construct a simple, pH-responsive molecular system for access control offers significant advantages.

In this work, we demonstrate a pH-regulated camouflage strategy. The core of this strategy is a triplex molecular switch that utilizes competitive binding sites to achieve programmable camouflage capabilities. We then integrated the switch into a scalable cascade logic circuit, constructing a two-layer DNA molecular firewall for access control. Governed by two distinct camouflage rules, the system can generate four unique operating modes. It effectively conceals the true input address and deceives unauthorized users by producing false outputs. All state switching is achieved simply by adjusting the external pH value. The strategy described in this work not only advances the field of molecular camouflage systems, but also provides a feasible and innovative approach to enhance the security of DNA cryptography.

2. Results and discussion

2.1. Design of a pH-responsive DNA triplex switch and demonstration of the camouflage mechanism principle

As a security control mechanism, firewalls restrict access according to predefined rules to prevent unauthorized entry and cyberattacks. Drawing inspiration from this concept and utilizing the pH-responsive behavior of the C–G–C+ triplex, we designed a pH-responsive DNA triplex switch capable of adopting distinct conformational states in two different environments. This allows it to operate as an effective DNA-based molecular firewall, with ON/OFF switching mediated by reversible changes in its molecular structure. The triplex switch comprises three modules: a true information storage module (Tr), a camouflage information storage module (Fl), and a switch module (S) (Fig. 1A). At pH 5.0, module S binds to module Tr through Hoogsteen bonds to form a triplex, setting the firewall to the F-ON state, in which only camouflaged access channels are exposed. At pH 9.0, S dissociates from Tr, switching the firewall to the F-OFF state and revealing the real access channels. To achieve a polymorphic response in the DNA triplex switch, we designed multiple structures with varying GC contents and analyzed triplex formation using polyacrylamide gel electrophoresis (PAGE). As shown in Fig. S1, the horizontal axis represents the GC content of each triplex. The results indicate that triplex-forming oligonucleotides (TFOs) with higher GC content require more H⁺ to form the CGC⁺ triplex structure, while a significant reduction in the required H⁺ is observed as the GC content decreases. Consequently, a triplex with a GC content of 33.3% was selected as the substrate for the DNA triplex switch. We verified the feasibility of the triplex switch through polyacrylamide gel electrophoresis (PAGE) experiments, with the results shown in Fig. 1B. Under acidic conditions, X in the control group bound to Tr and Fl to form a duplex structure (lane 6), whereas switch S in the experimental group successfully formed a DNA triplex structure with Tr and Fl (lane 5). In an alkaline environment, the substrates formed by both the control and experimental groups exhibited identical migration rates. These results effectively demonstrate that our triplex switch possesses excellent pH-responsive characteristics. Furthermore, to validate the pH-responsive behavior of the triplex switch, we examined the fluorescence intensities of the substrates under varying pH conditions. pH-insensitive fluorescent reporters (Cy5 and Cy3) were selected to ensure stable quenching efficiency (Fig. S2). The substrates displayed varying fluorescence intensities under different pH conditions (Fig. 1C and Fig. S3), indicating that as the pH increases (from 5.0 to 9.0), the DNA triplex gradually dissociates into a duplex structure. To further corroborate these findings, we replaced the triplex-forming domain (b**c**) in switch S with a random DNA sequence (f*). Under different pH conditions, this control system consistently showed high fluorescence intensity and was unable to form the C–G–C⁺ triplex (Fig. 1D and Fig. S4). These results further confirm that the triplex switch exhibits pronounced pH-responsive properties. Finally, to verify the pH reversibility of the switch, we performed a pH-cycling experiment by repeatedly adjusting the environmental pH between 5.0 and 9.0. We observed that the fluorescence intensity fluctuated periodically with the pH changes (Fig. 1E), providing convincing evidence for the excellent pH-responsive behavior and reversibility of the triplex switch under varying pH conditions. Therefore, the triplex switch can effectively drive state transitions of the molecular firewall through its pronounced pH responsiveness. The camouflage channel is the core component of the DNA molecular firewall for intrusion prevention. We constructed a pH-regulated camouflage mechanism. Initially, the firewall is maintained in an acidic environment. Attackers, unaware of the internal rules, will directly input the key after obtaining it. They can then only open the camouflage channel and access pre-set false information. Since the real information is not publicly available, attackers cannot determine the accuracy of the information obtained. In contrast, legitimate users, aware of the internal rules, can proactively switch the environment to alkaline conditions and then input the key to open the real channel, thereby obtaining the real information. We present a schematic diagram of the camouflage mechanism and its DNA reaction network (Fig. 2A). The camouflage capability is illustrated by the reaction process of key K with the DNA triplex substrate in two distinct states. To begin with, we performed polyacrylamide gel electrophoresis (PAGE) experiments to verify the relevant structures (Fig. 2B). Under acidic conditions, no DNA strand displacement reaction occurred between R and Fl. In contrast, under alkaline conditions, the displaced quencher strand was clearly visible (Lane 10), providing preliminary confirmation of the feasibility of the camouflage mechanism. Subsequently, to further validate the mechanistic feasibility, we conducted fluorescence detection experiments. Under acidic conditions, the substrate formed a triplex via Hoogsteen interactions, placing the triplex switch in the F-ON state. At this point, the addition of K only displaced Fl. Free Fl could not initiate strand displacement reactions with the downstream fluorescent reporter R via its toehold domain (c*), resulting in almost no detectable fluorescence signals. Under alkaline conditions, the triplex dissociated into a double-stranded structure, thereby transitioning the switch to the F-OFF state. Notably, to achieve camouflage, we designed the left and right toehold regions of the triplex with analogous structural domains, ensuring that the left toehold domain (a*) had a longer base length than the one on the right (a*⁻). Thus, upon the addition of K, it first underwent a DNA strand displacement reaction with the longer domain (a*), leading to the release of Tr and displacement of the quencher strand B, thereby yielding a high fluorescence signal (Fig. 2C). These results demonstrated that distinct fluorescence intensities could be obtained under different pH conditions, effectively realizing the camouflage function. To ensure reaction sensitivity, we optimized the concentration of K. Specifically, we optimized the reaction performance of the DNA network by varying the concentration of K (over a range of 0–400 nM) while maintaining fixed concentrations of F-ON, F-OFF, and R. As expected, real-time fluorescence kinetic curves showed that, under alkaline conditions, the reaction rate gradually increased alongside the increasing K concentration (Fig. 2D), while in contrast, under acidic conditions, no appreciable fluorescence signal was detected even with increasing K concentration (Fig. 2E). Therefore, the concentration of key K was finalized at 300 nM to guarantee the reaction efficiency of the DNA network.

Fig. 1 The pH-regulated DNA triplex switch. (A) Schematic diagram illustrating the principle of the triplex switch functioning as a firewall. (B) PAGE analysis of the triplex switch and the control group under different pH conditions. (C) Normalized fluorescence intensity of the triplex switch within its pH-responsive window. (D) Normalized fluorescence intensity of a pH-insensitive DNA substrate under different pH conditions. (E) Real-time fluorescence intensity changes of the triplex switch during pH cycling from 5.0 to 9.0. The buffer pH was adjusted using glacial acetic acid and sodium hydroxide. The concentration of both [F-ON] and [F-OFF] was 200 nM.

Fig. 2 Camouflage defense mechanism of the DNA molecular firewall. (A) Schematic diagram illustrating the principle and process of the camouflage mechanism. (B) PAGE analysis of the DNA network reaction process under different pH conditions. (C) Real-time fluorescence monitoring of the DNA network reaction process controlled by a triplex substrate. Real-time fluorescence detection of the reaction rates in the DNA network at different key concentrations under (D) pH 5.0 and (E) pH 9.0 environments. Experiments were conducted at 25 °C with [F-ON, F-OFF, and R] of 200 nM.

2.2. Strategy for logical operation functional conversion based on the camouflage mechanism

Our designed camouflage mechanism embeds an intrinsic firewall within the logic computation unit, ensuring that its logic gates expose their genuine computational core only when receiving the correct environmental signal. Building on this, we further construct two function-switching strategies for logic operations using a triplex switch. First, we develop a reconfigurable strategy for the dual-input logic unit, enabling it to dynamically toggle between AND and OR Boolean functions. Under acidic conditions, input strands IN₁ and IN₂ hybridize with F-ON; the released strand (be) then recognizes the (e*) domain of the downstream reaction unit B. This displaces the strand (e⁻ fm), and the subsequently released strand binds to R₁, displacing the quencher strand (m), thereby generating a high fluorescence signal. Thus, the output “1” was generated in the presence of only one input, executing the “OR” function (Fig. 3A, C and Fig. S5, S6). In an alkaline environment, the DNA triplex dissociated into a duplex, and the triplex switch transitioned to the F-OFF state, thereby functioning as a thresholding device. When input strands IN₁ or IN₂ were introduced, they first interacted with the (a*) domain featuring a longer anchoring region. The displaced strand (bc) could only interact with the downstream reaction unit A and failed to release the downstream quencher strand (m), resulting in a nearly undetectable fluorescence signal. The input signal exceeded the threshold only when both IN₁ and IN₂ were input simultaneously. At this point, the excess input strand bound to the short anchoring domain (a*⁻), triggering a downstream strand displacement reaction that displaced the quencher strand (m), thereby generating a high fluorescence signal. Thus, “1” was output only when both IN₁ and IN₂ were present, performing the “AND” function (Fig. 3B, D and Fig. S6, S7). These results demonstrate that the proposed strategy enables precise switching of the logic computing unit between “OR” and “AND” functions. To enhance the accuracy and robustness of the “AND” logic operation, we minimized operational leakage by adjusting the length of the anchoring domain (a*⁻) for F-ON and F-OFF (over the range of 3–6 nt) and the tail length of the binding domain (b) of input strands IN₁ and IN₂ (over the range of 0–6 nt). The logic operation exhibited optimal performance when the anchoring domain length was 4 nt and the tail of the binding domain (b) was 2 nt (Fig. 3E and Fig. S8 and Table S1). The performance of a DNA logic circuit depends not only on its Boolean logic fidelity and signal gain but also on its operational stability over time and across different environments. Accordingly, we investigated its reaction kinetics under varying pH conditions. As shown in Fig. S9, the circuit maintained a high output signal after 24 hours of operation in two distinct environments, demonstrating its high stability and strong adaptability to complex settings.

Fig. 3 Logic function switching strategy based on the firewall camouflage mechanism. (A) Operational mechanism of the “OR” logic computing strategy under pH 5.0 conditions. (B) Operational mechanism of the “AND” logic computing strategy under pH 9.0 conditions. (C) Real-time fluorescence detection of the switchable logic operation strategy under pH 5.0 and (D) pH 9.0 environments. (E) Real-time fluorescence detection of the reaction efficiency of the switchable logic operation strategy under pH 9.0 environment. [F-ON, F-OFF, IN1, IN2, A, B, and R1] = 200 nM.

Although this work validated the feasibility of the DNA logic circuit in buffer solution and demonstrated excellent 24-hour stability, challenges posed by complex sample matrices still need to be considered for practical environmental monitoring applications. Nucleases commonly found in real-world samples (such as soil extracts and biological fluids) may degrade the DNA strand, leading to circuit malfunction; changes in ionic strength can affect DNA hybridization kinetics and binding affinity, thereby altering threshold response behavior; furthermore, interfering molecules present in the sample may non-specifically bind to or competitively inhibit the logic output of the circuit. To address these challenges, future research can introduce various mitigation strategies. For example, chemical modifications can enhance the DNA strand's resistance to nucleases; encapsulation technologies (such as liposomes, polymer nanoparticles, or hydrogels) can isolate the circuit from the external environment while allowing small molecule targets to diffuse in; at the device design level, internal references or signal normalization strategies can be introduced to correct for the effects of ionic strength or nonspecific adsorption. These strategies have already shown promise in related DNA sensing fields and are expected to provide a feasible path for the transformation of this system into practical samples.

We further employ a triplex-switch-controlled camouflage mechanism to conceal more complex logic computing units. Based on the aforementioned OR-gate design and by embedding two firewalls with identical camouflage sites, we construct a three-input cascaded logic operation camouflage system, where the output of one OR gate serves as an input to the second OR gate. This three-input cascaded strategy required only two triplex switches and one fluorescent reporter to achieve complex logic transformation functions. In an acidic environment, the two switches adopted the F₁-ON and F₂-ON states. At this stage, input strands IN₃ and IN₄ could recognize the toehold domain (a₁*⁻) of F₁-ON, triggering the dissociation of strand (b₁e₁). The dissociated strand then acted as the input for F₂–ON, resulting in the detachment of strand (e₁–f₁), which bound to R₂ and displaced the quencher strand (f₁m₁). This yielded a high fluorescence signal. The direct input of IN₅ could also hybridize with F₂–ON, generating a high fluorescence signal (Fig. 4A, C and Fig. S10, S12). In an alkaline environment, both F₁-OFF and F₂-OFF functioned as thresholding devices. Here, only the simultaneous input of IN₃, IN₄, and IN₅ could induce the detachment of strand (e₁–f₁), leading to the detection of a high fluorescence signal (Fig. 4B, D and Fig. S11, S12).

Fig. 4 Cascaded logic function switching strategy based on the firewall camouflage mechanism. (A) Operational mechanism of the “OR–OR” logic computing strategy under pH 5.0 conditions. (B) Operational mechanism of the “AND–AND” logic computing strategy under pH 9.0 conditions. (C) Real-time fluorescence detection of switchable cascade logic operations in the pH 5.0 environment and (D) pH 9.0 environment. [F₁-ON, F₁-OFF, F₂-ON, F₂-OFF, R₂, and IN5] = 200 nM and [IN3 and IN4] = 250 nM.

2.3. Dual-layer DNA molecular firewall access control systems

Through the investigation of cascaded circuits, we have discovered that two firewalls can be integrated within the same system. Leveraging the structural flexibility of triplex-based camouflage sites, we have incorporated two distinct DNA triplex switches with different camouflage sites, enabling more complex expansion: one genuine site is exposed under acidic conditions, while the other is exposed under alkaline conditions.

This design establishes distinct decryption rules, effectively creating an inner and an outer firewall layer with differentiated access policies. Building on this, we have developed a pH-regulated dual-layer DNA molecular firewall access-control system. This system is capable of verifying legitimate network access while proactively misleading malicious attacks, thereby significantly enhancing defense-in-depth capability. The molecular system primarily comprises three modules: (1) a firewall molecular device module, (2) an information storage module, and (3) a system authorization module. Regarding the firewall molecular device module, it is designed primarily to resist malicious decryption attempts by external attackers and consists of both outer- and inner-layer DNA firewall molecular devices. The outer-layer device serves as the first line of defense against external attacks, verifying whether the user holds the correct Key 1 to determine access permission. Upon detecting the correct Key 1, the device grants preliminary access authorization. Subsequently, it validates the legitimacy of the access request in accordance with the pre-defined rule. Specifically, it opens the real channel for legitimate requests or provides a misleading camouflage channel for illegitimate ones. The inner-layer firewall molecular device, as the second line of defense, targets attackers who have breached the external protection. It employs differentiated internal rules to validate the legitimacy of access, thereby increasing the difficulty of decryption. For the information storage module, it stores the system's information set, including real information (OU₂ and OU₅) and fake information (OU₃ and OU₄). Only the combination of OU₂ and OU₅ represents the final correct information, while all other combinations represent incorrect or misleading information. Notably, we employed binary digits to represent output signals: a high fluorescence output (red for Cy3, blue for Cy5) corresponds to “1,” whereas a low fluorescence output corresponds to “0,” thus facilitating the differentiation of results. The system authorization module is primarily responsible for issuing keys for accessing the inner-layer firewall molecular device to users who have passed the first line of defense. To confuse attackers, even if fake information OU₃ is obtained through unauthorized access, Key 2 is still issued to mislead them (Fig. 5A and C).

Fig. 5 Double-layer DNA molecular secure access control system. (A) Conceptual diagram of the system. (B) DNA reaction network flowchart for user access to the molecular system. (C) Flowchart of user access to the system and a schematic diagram of obtaining different result information. (D)–(G) Real-time fluorescence detection results for user access to the molecular system via different methods. All substrates and inputs were used at 200 nM.

We used David as an example to illustrate the process of successful access to the molecular system. The dual-layer camouflage rules of the system adopted pH 5.0–9.0 as the valid decryption condition, with Tr₁ and Tr₂ serving as the two real access channels. The specific process went as follows: (1) Breaching the outer-layer firewall molecular device: David gained legitimate access to the outer system using the correct K₁. The DNA triplex in the outer system, initially stored in an alkaline environment, unwound into a duplex, activating the outer firewall molecular device. To access the real channel Tr₁, it was necessary to adjust the environmental pH to an acidic level in accordance with the rules; only under this condition could K₁ recognize the domain (a₂*⁻) and further gain access to Tr₁. (2) Information acquisition and system authorization: David leveraged Tr₁ to recognize the exposed toehold domain (e₂*) in the outer storage system, thereby acquiring OU₂. Upon receiving the request, the system revealed the module storing K₂. David then used OU₂ to recognize the anchor domain (e₂*⁻) in the system, obtaining K₂. (3) Decrypting the inner-layer firewall molecular device: David input the acquired K₂ into the inner system. Notably, the inner firewall molecular device followed different rules from the outer one: the environmental pH had to be adjusted to an alkaline state. After K₂ was input, it bound to the longer anchor domain (a₃*), causing Tr₂ to dissociate. David then input Tr₂ into the inner storage system where Tr₂ recognized the anchor domain (c₄*), ultimately obtaining OU₅ (Fig. 5B and F).

We also investigated the results when David adopted other illegal access methods: when he used the pH 5.0–pH 5.0 access method, the information obtained was OU₂ and OU₄ (Fig. 5D); when he used the pH 9.0–pH 9.0 access method, the information obtained was OU₃ and OU₅ (Fig. 5E); and when he used the pH 9.0–pH 5.0 access method, the information obtained was OU₃ and OU₄ (Fig. 5G). Combining all the information acquired via these illegal access methods yielded only misleading results. Finally, we validated the advantage of the dual-layer DNA molecular secure access control system in resisting brute-force attacks: when an attacker inputs an excessive number of K₁ keys in an alkaline environment, although the real access channel is opened, the camouflage access channel is also activated simultaneously. Thus, although the attacker obtained real information, they also obtained false information, and the final result of combining the two remained erroneous (Fig. S13).

To enhance system security, future work can be extended in two main directions: introducing molecular switches with distinct pH responses and increasing the number of firewall layers. As a proof of concept, we constructed an access control flowchart for dual switches operating at pH 6.0 and pH 8.0 (Fig. S14). By precisely modulating the pH-sensitive domains, the response window of each switch can be finely tuned, establishing the basis for an orthogonal switch library. In parallel, we developed a conceptual framework for a three-layer DNA firewall system (Fig. S15). Each additional layer with an orthogonal pH threshold doubles the number of possible state combinations (from 2ⁿ to 2ⁿ⁺¹): for instance, a three-layer system yields 8 states, a four-layer system 16 states, and so on, enabling exponential growth in security capacity.

In summary, we have successfully developed a molecular programming-based access control system featuring highly scalable outputs. The output chain, generated as the computational result, is expected to function as a universal adapter for diverse downstream applications—for example, serving as a “key” to unlock DNA nanostructures for triggered cargo release, reconstituting enzyme systems such as luciferase for signal amplification, or coupling with DNAzymes to produce colorimetric readouts suitable for instrument-free, on-site detection. These potential strategies highlight that our system not only enables molecular-level decision-making but also holds promise for translating such decisions into tangible outputs, offering new avenues for the design of intelligent molecular diagnostic platforms.

3. Conclusions

In this study, a pH-regulated molecular firewall access control system was constructed that strengthens the security and adaptability of DNA molecular information systems through a camouflage mechanism. The system core consists of a pH-responsive triplex switch that employs competitive inhibitory toeholds to disguise input addresses. This design enables the system to reconfigure its logical operations dynamically according to environmental signals and implement a sophisticated two-layer access control strategy. The camouflage mechanism serves as the primary defense of the molecular system against intrusion. By generating realistic false outputs in response to unauthorized access attempts—even those using correct keys—the system effectively misleads attackers and obfuscates genuine internal data. This mechanism ensures strong security against key leakage and brute-force attacks, as attackers cannot distinguish between authentic information and decoys, thereby significantly enhancing the system's stealth capabilities and robustness. Furthermore, this system can be integrated with various functional downstream operations, translating molecular authentication decisions into observable outputs. In summary, our strategy demonstrates the capability to construct intelligent, environmentally responsive barriers at the molecular level. By shifting the security paradigm from static key management to dynamic, rule-based access control, our work substantially reduces the key design complexity and overcomes the limitations of conventional management models. This method provides a simpler approach to constructing safe and adaptable molecular systems, with direct application potential in information security and environmental monitoring.

4. Methods

4.1. Materials and reagents

The oligonucleotides used in this study (Tables S2–S6) were designed using the NUPACK analysis module (https://www.nupack.org/). All DNA sequences were purchased from Shenggong Biotechnology (Shanghai) Co., Ltd. Unmodified DNA strands were purified by PAGE and ULTRAPAGE, whereas DNA strands modified with fluorescent dyes (Cy5 and Cy3) and a quencher (BHQ-2) were purified by high-performance liquid chromatography (HPLC). All DNA strands were dissolved in ultrapure water to prepare stock solutions. Quantification was performed using a Nanodrop 2000 spectrophotometer (Thermo Fisher Scientific Inc., Waltham, MA, USA), with absorbance measured at λ = 260 nm. Other chemicals were of analytical grade and used without further purification.

4.2. Native PAGE characterization

First, a 16% native polyacrylamide gel was prepared. The buffer used was 1 × TAE/Mg²⁺, adjusted to a pH of 9.0 or 5.0. Then, all samples were run in the 1 × TAE/Mg²⁺ buffer at the corresponding pH values at a constant voltage of 90 V for 120 minutes at 25 °C. Subsequently, the gel was stained with Stains-All solution for 30 minutes in the dark. Finally, after destaining under natural light, the gel was imaged using a scanner (CanoScan LIDE 120, Tokyo, Japan). For gel-based validation of the camouflage mechanism, the samples were first incubated at room temperature for 120 minutes after annealing, followed by PAGE analysis.

4.3. Fluorescence analysis

Fluorescence spectroscopic measurements were performed as follows: All fluorescence experiments were conducted using a microplate reader for detection in 1 × TAE/Mg²⁺ buffer at 25 °C. Cy5 fluorescence was detected at an excitation wavelength of 620 nm and an emission wavelength of 665 nm, whereas Cy3 fluorescence was detected at an excitation wavelength of 535 nm and an emission wavelength of 580 nm. In this experiment, all samples were prepared in a volume of 100 µL, with a detection interval of 1 minute per cycle. All experimental results are presented as the average of three replicate fluorescence measurements.

4.4. Fluorescence kinetic normalization method

The fluorescence minimum of the initial scan for each report sample (200 nM) was taken as the baseline. Then, fluorescence signals were collected once the input was added. Parallel experiments were conducted for logical operations with different inputs. For the given Cy5 and Cy3 fluorophores, experiments were completed to obtain at least one output signal (maximally open). The highest fluorescence increase among the reactions was normalized as “1”, and the other results were divided by this value to normalize the signals.

Author contributions

DeChun Tian: writing – original draft, validation, and methodology. Peijun Shi: visualization, conceptualization, and investigation. XiaoKang Zhang: methodology and investigation. Lijun Sun: software and formal analysis. Bin Wang: supervision and resources. Qiang Zhang: supervision, funding acquisition, and project administration.

Conflicts of interest

There are no conflicts to declare.

Data availability

All data supporting this article, including the raw data and images from fluorescence and PAGE experiments, are available on GitHub (https://github.com/tiandechun/Research-data).

Additional supporting data, such as DNA sequences, are included in the supplementary information (SI). Supplementary information is available. See DOI: https://doi.org/10.1039/d5nh00848d.

Acknowledgements

This work is supported by the 111 Center (No. D23006), the National Natural Science Foundation of China (No. 62272079, 62572088, and 62502063), the National Foreign Expert Project of China (No. D20240244), the Natural Science Foundation of Liaoning Province (No. 2024-MS-212 and 2024-BS-267), the Scientific Research Project of Liaoning Provincial Department of Education (No. LJ222411258005), the LiaoNing Revitalization Talent Program (No. XLYC2403039), the Artificial Intelligence Innovation Development Plan Project of Liaoning Province (No. 2023JH26/10300025), the Joint Plan of Liaoning Province Science and Technology Plan (No. 2024JH2/102600064 and 2024-MSLH-009), the Dalian Outstanding Young Science and Technology Talent Support Program (No. 2022RJ08), the Dalian Major Projects of Basic Research (No. 2023JJ11CG002), the Dalian Young Science and Technology Star Program (No. 2023RQ056), and the Interdisciplinary Project of Dalian University (No. DLUXK-2024-YB-001, DLUXK-2025-FX-003, DLUXK-2025-QN-022, and DLUXK-2024-QN-002).

Notes and references

1. D. J. Bernstein and T. Lange, Nature, 2017, 549, 188–194.
2. D. Joseph, R. Misoczki, M. Manzano, J. Tricot, F. D. Pinuaga, O. Lacombe, S. Leichenauer, J. Hidary, P. Venables and R. Hansen, Nature, 2022, 605, 237–243.
3. S. Agarwal and E. Franco, J. Am. Chem. Soc., 2019, 141, 7831–7841.
4. L. Wu, G. A. Wang and F. Li, J. Am. Chem. Soc., 2024, 146, 6516–6521.
5. L. Liu, Q. Hu, W. Zhang, W. Li, W. Zhang, Z. Ming, L. Li, N. Chen, H. Wang and X. Xiao, ACS Nano, 2021, 15, 11573–11584.
6. Y. Ma, Q. Wang, S. Du, J. Luo, X. Sun, B. Jia, J. Ge, J. Dong, S. Jiang and Z. Li, ACS Nano, 2024, 18, 5418–5433.
7. M. Yu, X. Tang, Z. Li, W. Wang, S. Wang, M. Li, Q. Yu, S. Xie, X. Zuo and C. Chen, Chem. Soc. Rev., 2024, 53, 4463–4489.
8. E. Benson, R. C. Marzo, J. Bath and A. J. Turberfield, Sci. Rob., 2022, 7, eabn5459.
9. Y. Li, Z. Gao, Y. Du, Y. Han, X. Ren, D. Wu, H. Ma, H. Ju, F. Xia and Q. Wei, Anal. Chem., 2025, 97, 8506–8515.
10. I. Qiqieh, J. Alzubi and O. Alzubi, Computing, 2025, 107, 35.
11. Z. P. Chen, W. J. Zeng, Y. M. Lei, W. B. Liang, X. Yang, R. Yuan, C. Yang and Y. Zhuo, Angew. Chem., Int. Ed., 2025, 64, e202414064.
12. J. Yin, S. Wang, J. Wang, Y. Zhang, C. Fan, J. Chao, Y. Gao and L. Wang, Nat. Mater., 2024, 23, 854–862.
13. Y. Zhang, B. Yan, X. Li, H. Liu, X. Liu, X. Xiao, Z. Mao and Z. Ming, Nanoscale, 2025, 17, 333–344.
14. X. Liu, Z. Chen, K. Wan, Y. Luo, J. Yang, L. Li, K. Tao, X. Xiao and M. Zhang, ACS Nano, 2025, 19, 9906–9914.
15. X. Wei, M. Ye, H. Jia, X. Zhou, Y. Wang, M. Li, C. Xue, L. Xu and Z. Shen, Biosens. Bioelectron., 2025, 271, 116975.
16. Y. Zhu, X. Zheng, R. Zhu, H. Zhao, H. Zhai, F. Qian, T. Zhang, Z. Xie, S. Liu and B. Jiang, Chem. Eng. J., 2023, 468, 143494.
17. K. J. Souza and D. K. Agrawal, Med-X, 2024, 2, 1.
18. R. R. Sinden, C. E. Pearson, V. N. Potaman and D. W. Ussery, Advances in genome biology, Elsevier, 1998, vol. 5, pp. 1–141.
19. W. Fu, Y. Zhao, J. Cao, X. Zhao, F. Chen, C. Fan and Y. Zhao, Adv. Mater., 2025, 2418688.
20. X. Zhang, Y. Liu, B. Wang, S. Zhou, P. Shi, B. Cao, Y. Zheng, Q. Zhang and N. Kirilov Kasabov, ACS Nano, 2023, 17, 18178–18189.
21. M. Hu, M. Yang, X. Cheng and T. Wu, Small, 2024, 20, 2400261.
22. A. Amodio, B. Zhao, A. Porchetta, A. Idili, M. Castronovo, C. Fan and F. Ricci, J. Am. Chem. Soc., 2014, 136, 16469–16472.
23. A. Idili, A. Vallée-Bélisle and F. Ricci, J. Am. Chem. Soc., 2014, 136, 5836–5839.
24. G. Felsenfeld, D. R. Davies and A. Rich, J. Am. Chem. Soc., 1957, 79, 2023–2024.
25. Y. Hu, A. Cecconello, A. Idili, F. Ricci and I. Willner, Angew. Chem., Int. Ed., 2017, 56, 15210–15233.
26. M. Dalla Pozza, A. Abdullrahman, C. J. Cardin, G. Gasser and J. P. Hall, Chem. Sci., 2022, 13, 10193–10215.
27. X. Sun, D. Yao and H. Liang, Nano Lett., 2023, 23, 11540–11547.
28. Y. Qin, J. Li, X. Zhang, K. Wan, H. Zhang, F. Huang, L. Wang, H. Wang, L. Li and X. Xiao, Chin. Chem. Lett., 2024, 35, 108826.
29. X. Fan, X. Zhang, Y. Zhang, S. Jiang and W. Song, Biosens. Bioelectron., 2024, 262, 116540.
30. Y. Zhou, L. Yang, X. Zhang, L. Zhu, X. Xiong, T. Xiao and L. Zhu, Anal. Chim. Acta, 2025, 1334, 343426.
31. Y. Guo, D. Yao, B. Zheng, X. Sun, X. Zhou, B. Wei, S. Xiao, M. He, C. Li and H. Liang, ACS Nano, 2020, 14, 8317–8327.
32. X. Xiong, M. Xiao, W. Lai, L. Li, C. Fan and H. Pei, Angew. Chem., Int. Ed., 2021, 60, 3397–3401.
33. Q. Tang, W. Lai, P. Wang, X. Xiong, M. Xiao, L. Li, C. Fan and H. Pei, Angew. Chem., Int. Ed., 2021, 60, 15013–15019.
34. P. Shi, X. Zhang, S. Cui, L. Sun, X. Liu, B. Wang and Q. Zhang, Chem. – Eur. J., 2025, 31, e202404541.

---

Footnote

† These authors contributed equally.

---