Vent^® (exo-) DNA Polymerase and Nt. BstNBI were purchased from New England Biolabs (NEB). Enhanced K-Blue TMB substrate was obtained from Neogen Corporation (United States). DNA marker (25–500 bp), deoxynucleotide triphosphates (dNTPs), tris (2-carboxyethyl) phosphine hydrochloride (TCEP), and streptavidin were supplied by Sangon Biotech Co., Ltd. (Shanghai, China). LMP-1 antibody-coated microplates and LMP-1 were acquired from Shanghai Fusheng Industrial Co., Ltd., while detection antibodies were sourced from Abcam. Biotin-labeled goat anti-rabbit IgG (second antibody) were obtained from Beyotime Biotechnology (Shanghai, China). Chloroauric acid and chloroplatinic acid were purchased from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). HPLC-purified DNA oligonucleotides were synthesized by Sangon Biotech Co., Ltd. (Shanghai, China), and the corresponding sequences are listed in Table 1. Ultrapure water (18.2 MΩ cm) was prepared using a Milli-Q system.

A mixture of ultrapure water (50 mL), chloroplatinic acid (0.38 mL, 10 mg mL^-1), and chloroauric acid (0.3 mL, 10 mg mL^-1) was prepared in a 100 mL flask and heated under stirring until boiling for 10 min. Subsequently, trisodium citrate (0.2 mL, 100 mg mL^-1) was introduced, and the reaction proceeded under reflux for 30 min. During this process, the solution gradually changed from nearly colorless to black, indicating the formation of AuPt alloy nanoparticles. Afterward, the resulting AuPt alloy nanoparticle dispersion was naturally cooled to room temperature and stored at 4 °C until use.

To conjugate the anti-digoxin antibody, the pH of the colloidal AuPt solution was adjusted to approximately 8.5 using 0.1 M potassium carbonate (K₂CO₃). A defined concentration of antibody (typically 10 μg mL^-1 in PBS) was slowly added to the AuPt solution under gentle stirring, followed by incubation at room temperature for 60 min to promote effective binding. After conjugation, bovine serum albumin (BSA, final concentration 1%) was added as a blocking agent to passivate the unmodified surfaces of the nanoparticles, and the mixture was incubated for an additional 30 min. The antibody-conjugated AuPt nanoparticles were then collected by centrifugation at 10,000 rpm for 10 min, and the resulting pellet was resuspended in PBS. The final antibody-labeled AuPt conjugates (Ab-AuPt) were stored at 4 °C until use.

Before modification, the gold electrode (AuE) underwent sequential polishing on a nylon pad with 0.5 μm and 0.3 μm alpha alumina powders, each step followed by thorough rinsing with double-distilled water. To eliminate potential organic contaminants, the electrode was immersed in a piranha solution (H₂SO₄/30% H₂O₂, 7:3 v/v) for 15 min and subsequently rinsed thoroughly with double-distilled water. Electrochemical activation of electrode interface was then performed in 0.5 M H₂SO₄ using cyclic voltammetry (scanning from −0.3 to 1.55 V) until a stable oxidation peak characteristic of gold was obtained.

Tetrahedral DNA nanostructures (TDNs) were assembled according to a previously reported method. Four single-stranded DNA oligonucleotides (strands A, B, C, and D) were custom-synthesized and HPLC-purified by Sangon Biotech (Shanghai, China). Among them, strands B, C, and D were modified at their 5′ends with disulfide-protected thiol (-SH) groups to enable surface immobilization via Au–S bonds. Each oligonucleotide was dissolved in TE buffer (pH 8.0) to prepare 50 μM stock solutions. To assemble TDNs, 1.0 μL of each oligonucleotide was mixed with 41 μL of TM buffer (20 mM Tris, 50 mM MgCl₂, pH 8.0) in a centrifuge tube, followed by the addition of 5 μL of freshly prepared TCEP solution (30 mM) to reduce the disulfide-protected thiols. The mixture was heated at 95 °C for 2 min and rapidly cooled to 4 °C within 30 s, yielding a 1.0 μM TDN stock solution. To modify the electrode, 5 μL of the TDN solution was dropped onto the cleaned AuE surface and incubated at 4 °C for 12 h to allow immobilization. After incubation, the electrode was rinsed with double-distilled water to remove unbound or nonspecifically adsorbed DNA. The resulting TDNs/AuE was stored in TM buffer at 4 °C until further use.

To detect LMP-1, 100 µL of LMP-1 at varying concentrations was added to LMP-1 antibody-coated microplates and incubated at 37 °C for 30 min. After incubation, the plates were washed three times with PBST. Subsequently, 100 µL of detection antibody was added to each well, followed by another incubation at 37 °C for 30 min. The washing step with PBST was repeated three times, and 100 µL of secondary antibody was added. The plates were incubated at 37 °C for another 30 min, and the same washing procedure was performed. Next, 100 µL of streptavidin (SA) solution was added and incubated for 20 min at 37 °C. After washing the microplates, 100 µL of biotinylated SP-HP (1 µM) was added and incubated for 20 min. Subsequently, 100 µL of the SDA reaction mixture was added to each well. The mixture consisted of 0.5×Nicking Endonuclease buffer (25 mM Tris-HNO₃, pH 7.9, 50 mM NaNO₃, 5 mM Mg(NO₃)₂, 0.5 mM dithiothreitol), 0.4 U µL^-1 Nt·BstNBI nicking enzyme, 0.05 U µL^-1 Vent (exo-) DNA polymerase, 1 × ThermoPol buffer (10 mM NaNO₃, 20 mM NH₄NO₃, 20 mM Tris-HNO₃, pH 8.8, 2 mM Mg(NO₃)₂, 0.1% Triton X-100), and 250 µM dNTPs. The reaction was incubated at 53°C for 45 min, followed by heat inactivation at 95°C for 10 min. A 5 µL aliquot of the reaction solution was applied to the TDNs/AuE and incubated for 30 min at room temperature (∼25°C). After rinsing the electrode, 5 µL of Ab-AuPt nanozymes was added and incubated for another 20 min. Finally, the electrode was immersed in TMB substrate solution to measure the electrochemical signal.

Electrochemical measurements were conducted on a CHI 660E electrochemical workstation using a three-electrode setup, comprising a gold electrode (AuE) as the working electrode, a platinum wire as the counter electrode, and an Ag/AgCl reference electrode. Amperometric detection was carried out in TMB substrate at an initial potential of 150 mV, with the reduction current recorded 200 s after the nanozymes-mediated redox reaction reached equilibrium.

A 12% polyacrylamide gel was used for electrophoresis with 1× TBE as the running buffer. DNA samples (10 µL each) were combined with 5× loading buffer and loaded into the assigned wells. The electrophoresis was conducted at a constant voltage of 130 V for 30 min. Gels were stained with GelRed for 15 min, followed by visualization and imaging using a gel documentation system.

The working principle of the amperometric immunoassay method for detecting LMP-1 using SDA-mediated nanozyme capture on DNA tetrahedra-based sensor interface is illustrated in Figure 1. The recognition of LMP-1 is achieved through a sandwich immunoreaction on a microplate. Biotinylated secondary antibodies are introduced into the immunocomplex to incorporate biotin, enabling the capture of SA on the microplate via biotin-SA binding. To enable isothermal signal amplification, we designed a biotinylated SP-HP probe, which can bind to the SA in microplate after immunorecognition. The sequence and secondary structure of SP-HP are shown in Figure 2A. The SP-HP probe consists of four distinct regions: a stem, a loop, the cleavage site for Nt. BstNBI nicking enzyme, and a template region for trigger DNA synthesis. The stem of SP-HP, with a 3′hydroxyl group, directly serves as a primer to activate the polymerization activity of Vent (exo-) DNA polymerase. This polymerase synthesizes a complete double-stranded cleavage site and a trigger DNA fragment. The nicking enzyme Nt. BstNBI introduces nicks at the cleavage site, and Vent (exo-) DNA polymerase, which possesses strand displacement activity, extends from the nick, displacing the downstream trigger DNA fragment. This process generates single-stranded, free trigger DNA fragments. The nicking and polymerization cycle continues, producing a large quantity of trigger DNA fragments. These trigger DNA fragments hybridize with hairpin probes on the TDNs, unfolding the hairpin structure and exposing terminal digoxin groups for binding with the Ab-AuPt nanozymes signal tag. Under the catalytic action of AuPt nanozymes, an electroreduction current of oxidized TMB is generated, providing a quantitative indication of LMP-1 concentration. Notably, the SDA strategy adopted in this study is a linear amplification process, as only a single nicking site is embedded in the SP-HP structure.

We first employed polyacrylamide gel electrophoresis (PAGE) analysis to validate the designed SP-HP’s ability to mediate strand displacement amplification (SDA) reaction. As depicted in Figure 2B, lane 1 represents the DNA markers, providing molecular weight references for comparison. Lane 2 shows the SP-HP alone, demonstrating its expected molecular size. When SP-HP was incubated with Vent (exo⁻) DNA polymerase and dNTPs, a marked increase in molecular weight was observed (lane 3). This indicates that SP-HP effectively acts as both a primer and a template, initiating the polymerase-mediated amplification process. The interaction between SP-HP and the polymerase highlights its dual functionality, which is crucial for amplification efficiency. To further investigate the SDA process, SP-HP was combined with Vent (exo-) DNA polymerase, Nt. BstNBI, and dNTPs. The results showed not only an increase in SP-HP’s molecular weight but also the emergence of a distinct, smaller, and highly intense band (lane 4). This band corresponds to the trigger DNA fragments generated through SDA. These findings confirm that SP-HP successfully mediates SDA, producing a substantial amount of trigger DNA fragments. This capability underscores the effectiveness of SP-HP in driving signal amplification for downstream electrochemical applications.

PAGE analysis was also employed to characterize the successful synthesis of TDNs. TDNs were assembled from equimolar amounts of four oligonucleotides: strands A, B, C, and D. As shown in Figure 2C, lane 1 represents the DNA markers, which serve as molecular weight references. Lane 2 displays strand A alone, showing its expected molecular size. In lane 3, the mixture of strand A and strand B forms a larger molecular complex, as indicated by the increased molecular weight. Similarly, lane 4 demonstrates the addition of strand C, which further increases the molecular weight of the DNA complex. Finally, when all four strands (A, B, C, and D) were combined, lane 5 reveals a distinct band corresponding to a high molecular weight DNA structure. In addition, atomic force microscope (AFM) imaging was used to directly observe the morphology of the assembled nanostructures. As shown in the 2D AFM image (Figure 2D), the DNA tetrahedra appear as uniformly dispersed, monodisperse nanoparticles. The corresponding 3D topographic image (Figure 2E) reveals a clear pyramid-like shape, which is consistent with the expected geometry of DNA tetrahedra. This result confirms that the TDNs were successfully assembled through the hybridization of the four oligonucleotides.

Due to the excellent catalytic activity of Pt and the bioconjugation-friendly properties of Au, AuPt alloy nanozymes were selected as signal amplification elements in this work. The morphology of the synthesized AuPt alloy nanozymes was characterized by transmission electron microscopy (TEM). As shown in Figure 3A, the low-magnification TEM image reveals that the AuPt alloy nanozymes are uniformly dispersed and exhibit a nearly spherical morphology. The inset in Figure 3A presents the particle size distribution histogram obtained by analyzing multiple nanoparticles using ImageJ software. The average diameter was calculated to be 6.13 ± 1.02 nm, indicating a narrow size distribution and good monodispersity. A higher-magnification image (Figure 3B) further confirms their well-defined structure. In addition, a high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM) image and the corresponding elemental mapping (Figure 3C) confirm the homogeneous distribution of both Au and Pt elements throughout the nanoparticles. These observations collectively verify the successful synthesis of the AuPt alloy nanozymes with a well-defined bimetallic structure. To confirm the crystalline structure and alloy formation of the synthesized AuPt nanozymes, X-ray diffraction (XRD) analysis was performed, and the results are presented in Figure 3D. The XRD pattern reveals distinct diffraction peaks at 32.08°, 33.26°, 37.34°, 38.74°, 54.94°, and 57.14°, which correspond to the (111), (200), and (220) planes of face-centered cubic (fcc) gold and platinum, respectively. The simultaneous presence of characteristic peaks from both Au and Pt confirms the formation of bimetallic AuPt alloy structures rather than physical mixtures.

To evaluate the peroxidase-like catalytic activity of the AuPt alloy nanozymes, UV–visible absorption spectroscopy was employed using TMB and H₂O₂ as the substrate system. As shown in Figure 3E, the TMB + H₂O₂ system alone exhibited negligible absorbance, indicating that spontaneous oxidation of TMB did not occur under the experimental conditions. However, upon addition of the AuPt alloy nanozymes, two distinct absorbance peaks appeared at 652 nm and 370 nm, which are characteristic of the oxidized form of TMB. The inset photograph in Figure 3E clearly illustrates the visual color change: the TMB + H₂O₂ solution remains colorless, while the addition of AuPt alloy nanozymes results in a pronounced blue coloration, further indicating effective catalytic oxidation of TMB. These results demonstrate that the synthesized AuPt alloy nanozymes possess intrinsic peroxidase-like activity. Furthermore, the catalytic affinity of the AuPt alloy nanozymes toward the substrates TMB and H₂O₂ was evaluated through steady-state kinetic analysis. As shown in Figures 3F,G, Michaelis–Menten curves were constructed, and the apparent Michaelis–Menten constants (Km) were calculated to be 0.18 mM for TMB and 0.08 mM for H₂O₂. These relatively low Km values indicate strong substrate affinities, underscoring the excellent catalytic performance of the AuPt nanozymes and their suitability for application in the developed electrochemical biosensing platform.

To highlight the catalytic superiority of our nanozyme, we systematically compared its substrate affinity with representative reported systems. Fe₃O₄ nanoparticles exhibited a Km of 0.098 mM toward TMB/H₂O₂ (Zhang et al., 2021), while V₂O₅ nanowires showed a Km of 0.12 mM (Niu et al., 2022). Graphene oxide functionalized with carboxyl groups (GO–COOH) presented a higher Km of 0.45 mM (Song et al., 2010), and platinum nanoparticles displayed a Km of 0.60 mM (Fan et al., 2011). MOF-based nanozymes such as Fe-MIL-88NH₂ and MIL-53(Fe) both demonstrated low Km values of 0.084 and 0.083 mM, respectively (Pargoletti and Gogotsi, 2025). In contrast, the AuPt alloy nanoparticles developed in this work exhibited Km values of 0.18 mM for TMB and 0.08 mM for H₂O₂, indicating balanced substrate affinity. These comparative data establish that our nanozyme not only rivals but also surpasses many reported systems, striking an optimal balance between affinity and catalytic efficiency, which is advantageous for sensitive biosensing applications.

To investigate the feasibility of the developed method, we utilized cyclic voltammetry (CV) to verify the successful assembly of TDNs on the surface of AuE. Figure 4A illustrates the CV behavior of the electroactive ion pair [Fe(CN)₆]^3-/^4- at the AuE surface, both before and after TDNs assembly. As shown in curve a, the bare AuE exhibits a pair of well-defined, reversible redox peaks with high peak currents, corresponding to the redox activity of [Fe(CN)₆]^3-/^4-. However, after the assembly of TDNs on the AuE (curve b), both the reversibility and peak currents of the redox peaks are significantly reduced, a phenomenon consistent with previous reports (Dong et al., 2015). This is attributed to the large size of the TDNs, which hinders the diffusion of [Fe(CN)₆]^3-/^4- from the solution to the electrode surface. Additionally, the negatively charged DNA in the TDNs repels the negatively charged [Fe(CN)₆]^3-/^4-, further impeding its electrochemical activity at the electrode. These results demonstrate the successful assembly of TDNs on the AuE surface.

To monitor the stepwise interfacial construction illustrated in Figure 1B, electrochemical impedance spectroscopy (EIS) was employed. As shown in Figure 4B, the bare AuE exhibited a very low charge transfer resistance (R_et, curve a), indicative of fast electron transfer at the unmodified surface. After immobilization of tetrahedral DNA nanostructures (TDNs), R_et increased significantly (curve b), suggesting the successful formation of a negatively charged DNA layer that hindered electron transfer. Upon incubation with the SDA amplification products, R_et decreased by approximately 50% (curve c), which may be attributed to hybridization-induced structural rearrangement, resulting in increased permeability. Further introduction of the Ab-AuPt conjugates led to a substantial reduction in R_et (curve d), likely due to the excellent electrical conductivity of the AuPt nanozymes, which facilitated interfacial electron transfer once captured on the electrode. On the other hand, the assembly density of TDNs on the AuE surface was calculated to be 9.7 × 10⁹ molecules cm^-2 according to chronoamperometry measurements (Steel et al., 1998).

To evaluate the feasibility of the developed sensor for quantitative detection of LMP-1, we utilized the I−t curve to monitor the electroreduction current of the TMB substrate catalyzed by AuPt alloy nanozymes. In this experiment, a sandwich immunoreaction was performed with and without the presence of LMP-1, followed by the SDA process. The reaction system in the microplate was then transferred to the surface of the TDNs/AuE electrode. After incubation with Ab-AuPt, the I−t curve test was conducted in the TMB substrate solution. As shown in Figure 4C, in the absence of LMP-1, only a small background current was observed (curve a). In contrast, the presence of LMP-1 led to a significant increase in the electroreduction current of TMB (curve b), resulting in a signal-to-background (S/B) ratio of 11.9. This result demonstrates that the developed sensor can effectively detect LMP-1. This phenomenon was further validated by CV in the TMB substrate solution. As shown in Figure 4D, in the absence of LMP-1, two pairs of well-defined redox peaks were observed, which are attributed to the typical two-electron redox process of TMB (curve a). In contrast, in the presence of LMP-1, a pair of asymmetric redox peaks characteristic of an electrocatalytic reaction was observed (curve b). Furthermore, to visually verify the capture of Ab-AuPt onto the electrode surface in response to the presence of LMP-1, a colorimetric assay was performed by immersing the sensing electrode into the TMB substrate solution. As shown in Figure 4E, no color change was observed in the absence of LMP-1, indicating that Ab-AuPt was not immobilized on the electrode. In contrast, a distinct blue coloration appeared in the presence of LMP-1, confirming the successful capture of Ab-AuPt on the electrode surface. The corresponding UV–vis absorption spectra further support this result clearly. These findings further support the specific recognition and signal output capability of the developed method.

To further verify the role of TDNs in enhancing detection performance, the hairpin probe embedded in the TDNs was replaced with a standalone hairpin probe (HP), and the resulting electrode was termed HP/AuE. As depicted in Figure 4F, when HP/AuE was used as the sensor electrode, a low background current was also observed in the absence of LMP-1, and the current increased significantly in its presence. However, the S/B ratio was only 5.2, substantially lower than the 11.9 observed with TDNs/AuE. These results confirm that TDNs effectively improve the sensor’s signal conversion capability and detection performance.

To ensure the highest detection performance, we systematically optimized several experimental parameters, including the incubation time of LMP-1 in the microplate, the duration of the SDA reaction, and the hybridization time between the trigger DNA and the hairpin probe on TDNs. The response signal was quantified by measuring the current output. First, the incubation time of LMP-1 in the microplate was examined. As shown in Figure 5A, the current signal increased as the incubation time was extended, reaching a stable value after 30 min. Extending the incubation time beyond 30 min did not yield further improvements, suggesting that the system had reached saturation. Thus, 30 min was determined to be the optimal incubation time. Next, the SDA reaction time was investigated to maximize signal generation. As illustrated in Figure 5B, the current signal grew steadily as the reaction proceeded, stabilizing after 45 min. This indicates that sufficient trigger DNA fragments were produced within this period. Prolonging the reaction beyond 45 min offered no additional benefits, confirming that the amplification had reached its limit. Therefore, the optimal SDA reaction time was set at 45 min. Lastly, the hybridization time of the trigger DNA with the hairpin probe on TDNs was optimized. Figure 5C shows that the current signal increased with hybridization time, leveling off after 30 min. This stabilization suggests that equilibrium hybridization was achieved, ensuring maximal signal output. Hence, 30 min was selected as the optimal hybridization time.

Under optimized experimental conditions, the analytical performance of the developed amperometric immunoassay method for LMP-1 detection was evaluated. As shown in Figure 6A, the absolute current signal increased progressively with rising LMP-1 concentrations in the range of 0.1–1,500 pg mL^-1, demonstrating the sensor’s responsiveness to varying target concentrations. In addition, Figure 6B illustrates a strong linear relationship between the current response (I) and the logarithm of LMP-1 concentration (lgC_LMP-1). The linear regression equation was determined to be: I = 988.01 + 723.74 lgC_LMP-1, with an excellent correlation coefficient of R ² = 0.9988. The detection limit (LOD) was calculated to be 47 fg mL^-1 based on a signal-to-noise ratio of 3. These results validate the sensor’s broad linear range, high sensitivity, and its capability for accurate quantification of LMP-1, even at ultra-low concentrations.

The specificity of the developed amperometric sensor for LMP-1 detection was tested against several non-target proteins, including prostate-specific antigen (PSA), carcinoembryonic antigen (CEA), and alpha-fetoprotein (AFP). LMP-1 was used at a concentration of 10 pg mL^-1, while the interfering proteins were tested at a much higher concentration of 1 ng mL^-1. As illustrated in Figure 7A, a significant current response was observed exclusively for LMP-1, whereas the signals generated by PSA, CEA, and AFP were negligible. This result highlights the sensor’s excellent selectivity for LMP-1 detection. To evaluate the reproducibility of the developed method, we measured the response signals of assays prepared both within the same batch and from different batches. For the same-batch assays, the relative standard deviation (RSD) of the response signals was 4.5% (Figure 7B). For assays fabricated across different batches, the RSD was 5.6% (Figure 7C). These results demonstrate the sensor’s reliable reproducibility, providing consistent performance across repeated experiments and between fabrication batches. To evaluate the storage stability of the fabricated sensors, TDNs/AuE electrodes were stored at 4 °C and tested over a period of 7 days using 10 pg mL^-1 of LMP-1. As shown in Figure 7D, the current signal exhibited a gradual and minimal decrease, with the signal retaining 99.5%, 98.4%, and 96.8% of the initial value on days 3, 5, and 7, respectively. These results demonstrate that the sensor retains good stability over a one-week period and is suitable for short-term storage and practical use.

The practical utility of the developed amperometric immunoassay method was validated by testing its performance in spiked fetal bovine serum (FBS), a complex biological matrix. To create test samples, LMP-1 was spiked into FBS at three concentrations: 1 pg mL^-1, 100 pg mL^-1, and 1,000 pg mL^-1. The prepared samples were then analyzed using the sensor under optimized conditions. As summarized in Table 2, the recoveries for the spiked concentrations ranged from 97.3% to 107.6%, highlighting the method’s accuracy and robustness in detecting LMP-1 within a challenging sample environment. In addition, these spiked samples were also analyzed using an ELISA kit for comparison. The results obtained by the ELISA kit were consistent with those of our method, further confirming the validity and precision of the developed sensor. These results confirm the sensor’s capability for reliable LMP-1 quantification in practical applications, paving the way for its potential use in clinical diagnostics and real-world sample analysis.