Tetrahedral DNA Framework–Bridged Aptasensing Method for Ultrasensitive Determination of Cardiac Troponin I

Abstract

The development of a sensitive and accurate method for the determination of cardiac troponin I (cTnI), as an ideal biomarker of acute myocardial infarction (AMI), is critical for the early diagnosis of AMI. In this work, a tetrahedral DNA framework (TDF)-bridged aptasensing method was proposed for the ultrasensitive determination of cTnI via electrogenerated chemiluminescence (ECL) and fluorescence imaging methods, in which a specific anti-cTnI aptamer-tagged tetrahedral DNA framework (TDF-aptamer) was designed as a capture probe by incorporating a specific aptamer against cTnI within the three-dimensional TDF. The TDF-aptamer was modified onto a gold electrode surface through the Au–S bond, providing a specifically captured platform for cTnI with a 3D binding space at the nanoscale. One microsized signal probe (Ab2-biotin•SA@MB-Ru1) was synthesized by using magnetic Fe₃O₄ as carriers to load a large amount of ruthenium complex and a specific cTnI antibody. After the sandwich assay among the modified TDF-aptamer, cTnI, and the signal probe, both ECL and fluorescence signals were recorded for the cTnI assay with desirable selectivity and high sensitivity (linear range of 5 to 80 pg/mL using ECL intensity detection and 0.5 to 500 pg/mL using ECL and fluorescence imaging methods). This work demonstrated that the combination of the specific TDF-aptamer and microsized Ab2-biotin•SA@MB-Ru1 within the imaging method can enhance the specific recognition and detection sensitivity for cTnI, which is promising for AMI diagnosis.

Introduction

Cardiovascular diseases are becoming major emergent health problems worldwide. Cardiac troponin I (cTnI) is well-known as an ideal biomarker for cardiac damage and has been established as the “gold standard” in clinical practice. Compared to healthy individuals, blood cTnI levels notably increase in patients with cardiovascular disease, ranging from a few pg/mL in asymptomatic subjects to over 10⁴ pg/mL in those with pathological conditions. , For example, cTnI concentration is around 0.001 μg/L in normal patients, increases to 0.01 μg/L in heart failure, and rises to 100 μg/L in myocardial infarction patients. It is critical to develop a sensitive and accurate method for the determination of cTnI throughout the diagnosis and treatment process of cardiovascular diseases.

Lots of methods have been developed for cTnI assay, such as fluorescence, electrochemical, electrogenerated chemiluminescence (ECL, electrochemiluminescence) , and other intelligent detection methods. , Among them, electrogenerated chemiluminescence is a type of luminescence that occurs at or near the surface of the electrode as a result of electrochemical reactions and chemiluminescence reactions. The ECL method is receiving increasing attention in the determination of cTnI because of its lower background, controllability, and sensitivity. ECL immunoassay has been commercially used for cTnI assay (for example, Elecsys cobas e systems, in the range of 0.16–25 ng/mL) by using 2.8 μm immune-magnetic beads (with captured antibodies immobilized on their surfaces) and Ru­(bpy)₃ ²⁺ derivative-labeled second antibodies, both specifically binding to cTnI. In the presence of the coreactant tri-n-propylamine (TPA or TPrA) and under an applied potential, Ru­(bpy)₃ ²⁺ labels emit ECL emission. However, it is limited in detecting low-expression cTnI below pg/mL. Consequently, the development of an ECL method for highly sensitive detection of cTnI is urgently needed.

The core factors for the detection sensitivity of biomarkers include the affinity of recognition molecules and the interfacial characteristics of the biosensor or biosensing platform. The biosensing interface, where the biosensing platform specifically reacts with the target biomarker and signal transduction processes, stands as the paramount determinant of overall biosensing performance. Notably, the rational engineering of such interfaces is indispensable for enhancing the efficiency of molecular recognition events. In recent years, aptamer-based biosensors or biosensing systems have garnered increasing attention, attributed to the aptamer’s inherent high affinity and selectivity toward target analytes, alongside its favorable characteristics of robust stability and cost-effectiveness. For example, the cTnI-specific Tro4 aptamer has been reported for cTnI assay with a low K _d value (270 pM). However, the intrinsic flexibility of aptamers renders them prone to nonspecific aggregation and entanglement on biosensing surfacesa phenomenon that significantly compromises target accessibility. To address this challenge, tetrahedral DNA framework (TDF), an artificially self-assembled nanostructure constructed via the precise base-pairing properties of nucleic acids, has emerged as a pivotal biomolecular scaffold for biosensing systems with the advantages including facile synthesis, exceptional site-specific orientation, robust structural rigidity, and the capability for high-precision multifunctionalization. − Several works have been reported in biosensing systems by employing the high programmability of TDF on the interface, leading to well-defined orientation, suitable density, and configuration of the framework. , Lin et al. found a 20-fold increase in the reaction rate by using about 12.6 nm TDF through programmable engineering of a biosensing interface with TDA. Wang et al. reported a single DNA molecule quantitation method on DNA tetrahedron-decorated substrates for both preventing nonspecific adsorption and sensitive recognition of single DNA molecules. Sun et al. reported an electrochemical dual-aptamer biosensor for cTnI analysis by using DNA NTH-based Tro4 aptamer probes, which provide an enhanced spatial dimension and accessibility for capturing cTnI. Inspired by the great performance of the TDF-based biosensing method, we are curious as to whether the integration of the TDF-based biosensing platform into the ECL method might be used to improve the analytical performance for cTnI.

To this end, we present a TDF-bridged aptasensing platform for the ultrasensitive determination of cTnI by coupling a specific anti-cTnI aptamer-tagged tetrahedral DNA framework (TDF-aptamer) as the capture probe and a ruthenium complex (Ru1) along with secondary antibody-coated magnetic beads (SA@MB) (Ab2-biotin•SA@MB-Ru1) as the signal probe. First, the TDF-aptamer was designed by incorporating a specific aptamer against cTnI within the three-dimensional tetrahedral DNA frameworks (Scheme A). One microsized signal probe (named Ab2-biotin•SA@MB-Ru1) was synthesized by modifying the ruthenium complex and secondary cTnI antibody onto the surface of magnetic beads (SA@MB, Scheme B). After the TDF-aptamer conjugates were assembled onto the gold electrode surface via Au–S bond formation, an aptasensing platform was constructed (Scheme C). Following the sandwich immunoassay among the modified TDF-aptamer, target, and signal probe, the sandwich magnetic immunocomplex (named TDF immunocomplex) was specifically formed on the electrode surface. Ru1 on the electrode generated a detectable fluorescence signal and ECL signal for quantification analysis. The characterization and analytical performance of the TDF-bridged aptasensing method for the cTnI assay were studied by using ECL intensity, FL imaging, and ECL imaging techniques. Furthermore, the application of this aptasensing method was examined for clinical human serum samples and compared with clinical results.

1. Schematic Diagrams for the Synthesis of (A) TDF-Aptamer and (B) Ab2-biotin•SA@MB-Ru1. (C) Schematic Diagram of TDF-Bridged Aptasensing Method for the Determination of cTnI.

Experimental Section

Materials and Apparatus

Bis­(2,2’-bipyridine)-(4’-methyl-[2,2’-bipyridine]-4-carboxylic acid) N-succinimidyl ester bis­(hexafluorophosphate) ((Ru­(bpy)₂(mcbpy-O-Su-ester)­(PF₆)₂, abbreviated as Ru1) was purchased from Sigma-Aldrich (USA). Streptavidin-functionalized magnetic beads (abbreviated as SA@MB, Φ = 1.0 μm, 10 mg/mL, Dynabeads M-270 Streptavidin) were obtained from Thermo Fisher Scientific (USA). Cardiac troponin I (cTnI, 24 kDa), and biotinylated capture cTnI antibody (Ab2-biotin) were provided by Shenzhen Lifotronic Technology Co., Ltd. (China). The reaction buffer containing 180 mM TPA was provided by Accucise Diagnostics Inc. (China). Carcinoembryonic antigen (CEA), immunoglobulin G (IgG), alpha-fetoprotein (AFP), and epithelial cell adhesion molecule (EpCAM) were purchased from Abcam (UK). All oligonucleotides (listed in Table S1) were obtained from Sangon Biotechnology Co., Ltd. (China). 10 mM phosphate-buffered saline (PBS1, 8.2 mM Na₂HPO₄, 1.8 mM KH₂PO₄, and 2.7 mM KCl, pH 7.40), 10 mM PBS2 (1.8 mM NaH₂PO₄, 8.2 mM Na₂HPO₄, 100 mM NaCl, and 5 mM MgCl₂, pH 7.40), and 10 mM phosphate-buffered solution (PB, 8.2 mM Na₂HPO₄ and 1.8 mM KH₂PO₄, pH 7.40) were used. All chemical reagents were of analytical grade and used as received. Millipore Milli-Q water (18.2 MΩ·cm) was used. Apparatus and other materials were listed in the Supporting Information.

Synthesis of Microsized Signal Probes

The microsized signal probe (Ab2-biotin•SA@MB-Ru1, Probe A) was synthesized by first covalently coupling Ru1 onto SA@MB and then affinity binding Ab2-biotin onto SA@MB. Briefly, 100 μL of 10 mg/mL SA@MB suspension, 100 μL of Ru1 solution (10 mg/mL), and 500 μL of 10 mM PBS1 were mixed and incubated at room temperature for 12 h with slow tilting rotation. After magnetic separation and washing with 10 mM PBS1, the SA@MB-Ru1 was resuspended in 500 μL of 10 mM PBS1. Then, 50 μL of 1 mg/mL Ab2-biotin was added to the SA@MB-Ru1 suspension and incubated at 37 °C for 60 min. After magnetic separation and washing with 10 mM PBS1, Ab2-biotin•SA@MB-Ru1 was resuspended in 10 mM PBS1 (pH 7.40) at a concentration of 1 mg/mL, ready for use.

For comparison, another microsized signal probe (Ru1-Ab2-biotin•SA@MB, Probe B) was synthesized by first covalently coupling Ru1 with Ab2-biotin and then affinity binding Ru1-Ab2-biotin on SA@MB. The detailed process is listed in the Supporting Information.

Synthesis of TDF-Aptamer

The TDF-aptamer was synthesized according to the reference. First, ssDNA-1, ssDNA-2, and ssDNA-3 were reduced using a 10 mM TCEP solution. Second, 20 μL of 10 μM ssDNA-1, 20 μL of 10 μM ssDNA-2, 20 μL of 10 μM ssDNA-3, 20 μL of 10 μM ssDNA-4, and 120 μL of 10 mM PBS2 were mixed, incubated at 95 °C for 15 min, and then immediately transferred to 4 °C for annealing to form the TDF-apatmer. The concentration of the synthesized TDF-aptamer is considered to be 1 μM, and it is stored at 4 °C.

Fabrication of Aptasensing Platform

First, the gold electrode was cleaned using the conventional process. Second, 10 μL of 50 nM TDF-aptamer was dropped onto the surface of the gold electrode for 3 h at room temperature. After washing with 10 mM PB, the resulting electrode was dipped into 100 μL of 1 mM MCH for blocking for 30 min. After being washed with 10 mM PB, the aptasensing platform was formed.

Measurement of cTnI

For the sandwich assay of cTnI, the TDF-aptamer-modified electrode was first dipped into 100 μL of cTnI or samples and reacted at 37 °C in a water bath for 30 min. Then, the resulting electrode was dipped into 100 μL of 5 μg/mL Ab2-biotin•SA@MB-Ru1 signal probe at 37 °C for another 30 min. The human serum samples were provided by Shaanxi Normal University Hospital and diluted 10 times with 10 mM PBS before use.

Three detection techniques, including ECL intensity, FL imaging, and ECL imaging, were used for the determination of cTnI. For ECL intensity detection (mode I), an MPI-E ECL detector (Xi’an Remax Analysis Instruments Co. Ltd., China) was used as the detector. The obtained modified working electrode, along with an Ag/AgCl reference electrode and the platinum counter electrode, was placed in 0.1 M PBS1 containing 180 mM TPA (pH = 7.40). Step pulse (initial voltage is 0 V, and the duration is 2 s; the termination voltage is +1.4 V, and the duration is 10 s) was used for ECL measurement. The high voltage of the photomultiplier tube (PMT) was set at −900 V during the measurement. The maximum ECL intensity was used as the analytical signal for cTnI.

For ECL imaging and FL imaging, an IX73 inverted microscope equipped with a 60X objective and Prime 95B complementary metal-oxide-semiconductor (CMOS) was used as a detector. For FL imaging, FL images were collected in the 593–667 nm emission wavelength range with a 455–495 nm excitation wavelength for 200 ms of exposure time (mode II). ECL imaging was performed in 180 mM TPA by using the multipotential steps technique (an initial potential of 0 V for 0.2 s, a high potential of +1.4 V for 0.5 s, and an exposure time of 10 s, mode III). The obtained FL and ECL images were analyzed by using ImageJ software. A threshold in each FL and ECL image was determined as five standard deviations above the mean of the gray value of the background for identifying bright dots and counting the number of bright dots (bright particles) in FL and ECL images. The contour of each MB was recognized, and the amount of MBs was counted in the images using the ImageJ method. The number of bright particles in FL and ECL images was used as an analytical signal for the cTnI assay.

Results and Discussion

Synthesis and Characterization of TDF-Aptamer

In this work, one aptamer-tagged tetrahedral DNA framework (TDF-aptamer) was designed and synthesized as a capture probe for the cTnI assay. The TDF-aptamer was self-assembled via one-step hybridization among four ssDNA strands (Table S1, Figure A), where three ssDNA strands (ssDNA-1, ssDNA-2, ssDNA-3) contained thiol groups with 55 bases, and ssDNA-4 contained an exposed specific anti-cTnI aptamer with 95 bases. The formed TDF-aptamer contained 17 base pairs per edge, and three thiols (−SH) along with an anti-cTnI aptamer were located at the vertices of the TDF-aptamer (Figure S1). Theoretically, the height and the edge length of the TDF- aptamer were calculated to be 4.7 nm and 5.8 nm, respectively, due to the characteristic spacing of 0.34 nm between adjacent base pairs in a DNA double helix.

1.

Successful synthesis of the TDF-aptamer. (A) Schematic diagram of the TDF-aptamer. (B) PAGE analysis. Lanes 1–5 represent the DNA ladder marker, ssDNA-1, ssDNA-1+ssDNA-2, ssDNA-1+ssDNA-2+ssDNA-3, and ssDNA-1+ssDNA-2+ssDNA-3+ssDNA-4. DNA concentration is 0.2 μM. (C) TEM of the TDF-aptamer, the red triangle indicates the morphology and size of the TDF-aptamer. Scale bars are 20 nm. (D) Distribution of the edge length of TDF-aptamer.

In order to demonstrate that the TDF-aptamer was successfully synthesized, the obtained TDF-aptamer was characterized by polyacrylamide gel electrophoresis (PAGE) and transmission electron microscope (TEM). From the PAGE results in Figure B, it can be seen that a series of sharp, discrete migrated bands in the range of 50 to 200 bp were clearly visible for the DNA ladder marker. For ssDNA1, a sharp migrated band at 50 bp was observed. For the hybridized complex of ssDNA-1 and ssDNA-2, two sharp migrated bands at approximately 50 bp and 100 bp were seen. For the hybridized complex of ssDNA-1, ssDNA-2, and ssDNA-3, two main migrated bands at 50 bp and 150 bp were clearly visible. For the TDF-aptamer, a main migrated band above 200 bp was observed Second, the obtained TDF-aptamer was characterized using TEM to visualize its morphology. The obtained TEM images in Figure C showed numerous black dots attributed to the TDF-aptamer (circled with red triangles) that were uniform in size and well-dispersed across the field of view. From the enlarged inset, the morphology of the TDF-aptamer was triangular. In addition, the edge length of the TDF-aptamer in the TEM image was manually measured using ImageJ software and further analyzed. The obtained results (Figure D) showed that the edge length of the TDF-aptamer was approximately 6.0 nm, which is close to the theoretical value. These results indicated that the hybridization reaction of the four ssDNA strands indeed occurred and that the TDF-aptamer was successfully synthesized.

Synthesis and Characterization of Signal Probe Ab2-biotin•SA@MB-Ru1

In order to improve the sensitivity of the aptasensing method and to get good resolution for imaging, 1 μm streptavidin-coated magnetic beads were used as carriers to prepare the signal probe by immobilizing Ab2-biotin and Ru1 on them. Two microsized signal probes, including Probe A (Ab2-biotin•SA@MB-Ru1) and Probe B (Ru1-Ab2-biotin•SA) were synthesized and compared. Probe A was synthesized by first conjugating the ruthenium complex (Ru1) directly to the surface of magnetic SA@MB via a covalent coupling mode, followed by the specific coupling of Ab2-biotin on SA@MB (as shown in Figure S2A). Probe B was synthesized by first covalently coupling Ru1 with Ab2-biotin and then performing affinity binding of Ru1-Ab2-biotin on SA@MB (Figure S2B). The two obtained signal probes (Probe A and Probe B) were characterized using inductively coupled plasma optical emission spectroscopy (ICP-OES), UV–vis absorption spectroscopy, and fluorescence (FL) techniques.

First, the Ru amount within two signal probes (Probe A and Probe B) was determined by ICP-OES. The result (Table S2) showed that the amount of Ru was 3.18 fg per bead and 2.62 fg per bead for Probe A and Probe B, respectively, similar to that in ref. . Second, the two signal probes were imaged by using a confocal laser scanning microscope (CLSM). Bright FL emission could be observed on both Probe A and Probe B. Along the Z-axis CLSM within the range of 0 to 1.8 μm, it can be seen that Ru1 is located on the surface of MBs (Figures S3–S4). In order to confirm whether there is noncovalently associated Ru1 within Probe A and Probe B, the FL and ECL intensities of different samples during the washing process were recorded. The FL intensities of ten supernatants from both Probe A and Probe B after washing and magnetic separation nearly remained constant (relative standard deviation, RSD, is 7.8% for Probe A and 4.2% for Probe B, Figure S5). Additionally, the ECL intensities of Probe A and Probe B after washing and magnetic separation also nearly remained constant over ten cycles (RSD is 4.6% for Probe A and 10% for Probe B, Figure S6). This result indicated that no free leached Ru1 was detectable from the supernatants during the washing process and that Ru1 is stably loaded on the SA@MB.

To demonstrate the successful binding of Ab2-biotin in the signal probe, the immunocomplex (TDF-aptamer/cTnI/Probe A, TDF-aptamer/cTnI/Probe B) was formed at the gold electrode under different conditions and was imaged by FL imaging and ECL imaging. In the absence of cTnI, virtually no bright dots were observed in the FL images and ECL images for both Probe A and Probe B (Figure S7). In the presence of cTnI, numerous bright dots were simultaneously seen in the FL images (Figure S8A ₁ ) and the ECL image for Probe A (Figure S8A ₂ ). The corresponding pseudocolored images (Figure S8A ₃ ) demonstrated that the signals of bright dots in the ECL image were arising from the same bright dots in the FL image. For Probe B, some bright dots were seen in the FL image (Figure S8B ₁ ) and the merged image (Figure S8B ₃ ) while no bright dots were seen in the ECL image (Figure S8B ₂ ). By further statistical analysis of the average gray value in 40 regions of interest (ROI) in the FL and ECL images, the gray value from Probe A in both the FL and ECL images was higher than that from Probe B (Figure S8A ₄-B₄). These results indicated that Ab2-biotin was successfully labeled in both Probe A and Probe B, and the FL and ECL properties of Probe A are significantly better than those of Probe B.

In the MB-based ECL process, only the coreactant could be electrochemically oxidized to generate free radicals, which then react with Ru1 on the beads to emit light. The maximum distance at which light is observed is ∼6 μm, based on the half-life of the cation radical, which would be ∼0.2 ms. We previously found that both the cation radical lifetime and the optical propagation across MBs affect ECL emission. Although Probe B theoretically allows Ru1 to be closer to the electrode surface, while Probe A theoretically allows Ru1 to be farther away from the electrode, higher ECL emission was obtained by using Probe A. Third, we employed the Bradford protein assay in conjunction with a BSA calibration curve (Figure S9) to estimate the amount of Ab2-biotin conjugated to the magnetic beads in Probe A. An estimate of Ab2-biotin per bead was estimated to be 10795, similar to that in ref. . Considering the higher ECL emission, Probe A (Ab2-biotin•SA@MB-Ru1) was selected as the signal probe for the following experiments.

ECL Aptasensing Platform for the Determination of cTnI

Inspired by the good properties of the captured probe (TDF-aptamer) and the signal probe (Ab2-biotin•SA@MB-Ru1), an ECL aptasensing method was developed for the determination of cTnI. As shown in Scheme C, after the formation of the TDF immunocomplex (TDF-aptamer/cTnI/Ab2-biotin•SA@MB-Ru1) on the gold electrode, the ECL signal was recorded in the presence of TPA using an MPI-E ECL analyzer.

First, the TDF-aptamer was modified onto a gold electrode to form an aptasensing platform. To study the interaction between the TDF-aptamer and the Au electrode, the TDF-aptamer-modified gold foil was characterized using X-ray photoelectron spectroscopy (XPS). XPS analysis of Au revealed a characteristic doublet in the Au 4f spectrum, with the Au 4f₇/₂ and Au 4f₅/₂ components centered at 88.1 eV and 84.4 eV, respectively (Figure S10A). In contrast, the S 2p XPS signal appears broad and complex, requiring deconvolution into four distinct peaks for accurate fitting (Figure S10B). A prominent S 2p₃/₂ peak centered at 162.1 eVan iconic signature of sulfur involved in Au–S bondingconfirms the successful assembly of a thiol-functionalized TDF-aptamer on the gold electrode surface. Furthermore, the distribution of the TDF-aptamer on the gold foil was visualized using atomic force microscopy (AFM). The AFM images showed that the number of bright dots for the TDF-aptamer modified gold foil (Figure S11B) was significantly greater than that for bare gold foil (Figure S11A), indicating that the TDF-aptamer is indeed immobilized on the surface of the gold electrode.

The feasibility of the ECL aptasensing method for detecting cTnI was studied using PMT as a detector (mode I), where the peak ECL intensity (I) was used as an analytical signal. In the absence of cTnI (0 pg/mL), a negligible ECL signal (859 a.u.) at +1.4 V was observed (Figure S12). In the presence of cTnI (50 pg/mL), an obvious ECL emission (11709 a.u.) at +1.4 V was observed (Figure S12). This result indicates that the ECL aptasensing method for detecting cTnI using PMT as a detector (mode I) is feasible. The influence of the applied potential on ECL intensity was studied. As a case, 80 pg/mL of cTnI was selected. The result (Figure S13) showed that the ECL intensity increased with the applied potential from +0.9 V to a maximum at +1.4 V (18266 a.u.) and then decreased at higher potentials. Therefore, the applied potential of +1.4 V was selected to trigger ECL emission in the experiment.

Under optimal conditions, the developed ECL aptasensing method (mode I) was utilized for the cTnI assay using Probe A (Figure A). Figure C shows the ECL intensity versus time profiles obtained at +1.4 V for different cTnI concentrations. It could be seen that the ECL intensity gradually increased with increasing cTnI concentration. A good linear relationship was obtained between I and cTnI concentration in the range from 5 to 80 pg/mL (Figure E). The linear equation was I = 156.8c_cTnI + 3669 (the unit of c_cTnI is pg/mL, R ² = 0.9625), and the detection limit (DL) was calculated to be 1.2 pg/mL by the equation DL = 3σ_b/s (σ_b = standard deviation of blank (I); s = the slope of the calibration curve).

2.

(A–B) Schematic diagram of aptasening method for cTnI assay using Ab2-biotin•SA@MB-Ru1 (A) and Ru1-Ab2 (B) as signal probe. (C, E) ECL intensity-time curves (C) and the corresponding calibration plot (E) for cTnI by using Ab2-biotin•SA@MB-Ru1 as the ECL signal probe. (D, F) ECL intensity-time curves (D) and the corresponding calibration plot (F) for cTnI by using Ab2-Ru1 as the ECL signal probe. +1.4 V of applied potential.

For comparison, another ECL aptasensing method was also performed for the cTnI assay (Figure S14), where the signal probe Ab2-Ru1, synthesized by binding Ru1 to the Ab2, was utilized to replace Ab2-biotin•SA@MB-Ru1 (Figure B). The Ab2-Ru1 was verified by UV–vis absorption spectroscopy, and the molar ratio of Ru1:Ab2 within Ab2-Ru1 was calculated to be 1.48 by measuring the absorbance at 280 nm for Ab2 and 457 nm for Ru1 (Figure S15). The results in Figure D showed that the ECL intensity gradually increased with increasing cTnI concentration. A good linear relationship was obtained between I and cTnI concentration in the range from 0.1 to 1 ng/mL (Figure F). The linear equation was I = 14334c_cTnI + 1127 (the unit of c_cTnI is ng/mL, R² = 0.9976), and the DL was calculated to be 0.05 ng/mL. Thus, the good analytical performance of the ECL aptasensing method (mode I) using Ab2-biotin•SA@MB-Ru1 is better than that of the ECL aptasensing method using Ab2-Ru1. The selectivity of the developed aptasensing method was verified for different proteins such as AFP, CEA, EpCAM, IL-6, and IgG (Figure S16). In short, the developed ECL aptasensing method using PMT as a detector (mode I) has great analytical performance for the cTnI assay.

In addition, method reproducibility and sensor stability were studied. Different fabricated aptasensing biosensors were utilized for the determination of cTnI (0, 20, and 50 pg/mL) over four consecutive days to test sensor stability. The results (Figure S17) showed that the ECL intensities (I) for the blank were 1064 a.u. on the first day, 875 a.u. on the second day, 1009 a.u. on the third day, and 983 a.u. on the fourth day, respectively. The variation in ECL intensities across the four consecutive days fell within an acceptable margin of error (RSD = 8.1% for the blank, 12% for 20 pg/mL, and 12% for 50 pg/mL). The method reproducibility was tested by recording the ECL intensity for 20 pg/mL cTnI by using five different biosensors. The RSD was calculated to be 4.7%. Additionally, to investigate the stability of the synthesized TDF-aptamer, the prepared TDF-aptamer was stored at 4 °C over 25 days and then used to fabricate the aptasensing platform for the cTnI assay (0 and 20 pg/mL). On the first day of storage, the ECL intensities for 0 and 20 pg/mL cTnI were 1198 a.u. and 8198 a.u., respectively. In the range of 1 to 10 days, the ECL intensity decreased slightly. On the 10th day, the ECL intensities for 0 and 20 pg/mL cTnI were 874 a.u. and 7206 a.u., respectively. Beyond 10 days, the ECL intensity greatly decreased with the increase in storage time of TDF (Figure S18). This result indicated that most of the TDF in the solution may have dissociated. Therefore, the optimal storage time for the prepared TDF is within 10 days.

Feasibility of Dual ECL and FL Imaging of cTnI

Despite the good performance of the aptasensing platform for the cTnI assay using mode I, the ECL methods detected biomarkers at the ensemble level, where antigens bind with antibodies labeled with signal probes, which are limited for the detection of low-expression biomarkers at the early stage of the disease. In this work, we aimed to develop dual-mode digital FL (mode II) and ECL imaging methods (mode III) for the sensitive detection of cTnI, where CMOS was utilized as the detector.

First, an FL imaging method was investigated to validate the reliability of the digital quantitative method by using a TDF-bridged biosensing platform and a microsized signal probe. The configuration of FL imaging is shown in Figure A. For the digital imaging method, it is important to spatially isolate and discriminate single labels originating from single-molecule reaction events. In this work, the TDF-aptamer was employed as a biosensing platform to orientally capture the target on the electrode surface. To visualize the distribution of the TDF immunocomplex on the surface of the gold electrode, the formed TDF immunocomplex at the gold electrode for 100 pg/mL cTnI was recorded using FL imaging. For comparison, Probe A was also randomly dispersed on the surface of the gold electrode and recorded with FL imaging. The FL images showed that the distribution of bright dots for the TDF immunocomplex was more uniform than that for the signal probe alone (Figure S19). Besides, a metric named QD₉ was proposed for evaluating the uniformity of the formed TDF immunocomplex at the gold electrode in FL images (detailed description in the SI). The results showed that the QD₉ was calculated to be 37.3% for the TDF immunocomplex, while it was 60.2% for the signal probe alone, indicating that the TDF-aptamer can improve the uniformity of the random dispersion of the signal probe on the surface of the gold electrode, which is beneficial for counting the number of TDF immunocomplexes.

3.

(A) Schematic diagram of the configuration of FL imaging. (B) FL images for the detection of varying concentrations of cTnI (B₁–B₄: 1,100, 200, and 500 pg/mL). Scale bar: 30 μm. FL imaging conditions: 455–495 nm excitation wavelength, 593–667 nm emission wavelength, and 200 ms exposure time. (C) Number of bright dots vs cTnI concentration.

It should be noted that the SA@MB used in this work has autofluorescence emission under a 455–495 nm excitation wavelength and a 593–667 nm emission wavelength for 200 ms of exposure time. Under the same FL imaging conditions, Probe A emits much stronger fluorescence emissions (Figure S20B). However, compared to the gray value, SA@MB alone displays low and vague FL emissions (Figure S20A), with an average gray value centered at 330 a.u. (Figure S20C) while Probe A shows brighter and clearer FL emissions with an average gray value centered at 3400 a.u. (Figure S20D). In our work, a threshold in each FL image was determined as five standard deviations above the mean of the gray value of the background (608 a.u.) for identifying bright dots and counting the number of bright particles (bright dots). The effect of autofluorescence emission from SA@MB can be neglected.

Theoretically, to achieve all signals of the TDF immunocomplex formed at the gold electrode, about 81 FL images (or ECL images) were required to be obtained in one measurement due to the detector area of 0.048 mm² and the electrode area of 3.2 mm². However, this process is time-consuming. To validate whether a limited number of fields of view can reliably represent the entire electrode, we performed the following comparative analysis. Thirty images and 5 images were acquired from randomly distributed positions across the entire electrode surface. The number of bright dots from both sampling strategies was compared using a t test. The absence of a statistically significant difference (p > 0.05) confirms that the sampling protocol of 5 images yields representative data for the entire electrode surface (Figure S21). To shorten the recording time, 5 FL images were collected in 5 regions of the electrode (Figure S22) to represent all FL images on the entire gold electrode. By multiplying the average number of bright dots by 81, the integral number of bright dots was determined as the analytical signal. For example, for 100 pg/mL cTnI, the number of bright dots in 5 FL images obtained in 5 regions was 60, 72, 64, 74, and 65, respectively. Then, the average number of bright dots was 67, while the integral number of bright dots was calculated to be 5427.

Figure A shows FL images of the aptasensing platform for the cTnI assay (mode II) by using 5 μg/mL Probe A. Bright dots were observed against a black background in all FL images, and the number of bright dots increased with the rise in cTnI concentration (Figure B; detailed images are listed in Figure S23A). The number of bright particles obtained from the FL images shows a certain trend (Figure C). Based on the aforementioned quantification method, the number of fluorescent bright dots exhibited a good linear relationship with the cTnI concentration within two ranges: 0.5 pg/mL to 10 pg/mL (Figure S23B) and 20 pg/mL to 500 pg/mL (Figure S23C). Linear regression analysis yielded the equations N = 312.6C + 967.7 (R ² = 0.9623) and N = 22.6C + 3284 (R ² = 0.9903), respectively, and the DL was 0.5 pg/mL. By increasing the amount of Probe A to 10 μg/mL, the number of fluorescent bright dots exhibited a good linear relationship with the cTnI concentration within two ranges: 1 pg/mL to 50 pg/mL and 50 pg/mL to 500 pg/mL (Figure S24). Linear regression analysis yielded the equations N = 87.4C + 2620.2 (R = 0.8920) and N = 22.3C + 6262.4 (R ² = 0.9579), respectively, and the DL was 1 pg/mL. Different analytical performance was obtained due to varying experimental conditions.

Furthermore, the ECL imaging technique was further used to tset the feasibility of the digital ECL imaging method for the cTnI assay. The configuration of ECL imaging is shown in Figure A. Due to the oxidation of the thiol group up to a potential of +0.8 V, there is no obvious ECL emission upon the successive application of a potential of +1.4 V more than once. Only one image was recorded for ECL imaging. Figure B shows ECL images for different concentrations of cTnI (detailed images are provided in Figure S25). The particle number obtained from the ECL images showed an increasing trend with the rising concentration of cTnI from 0.5 to 500 pg/mL (Figure C), indicating that the designed ECL imaging method is capable of detecting cTnI. The number of ECL bright dots exhibited a good linear relationship with the cTnI concentration within two ranges: 0.5 to 8 pg/mL (Figure S25B) and 10 to 500 pg/mL (Figure S25C). Linear regression analysis yielded the equations N = 3.8C + 12.3 (R ² = 0.9305) and N = 0.2C + 45.9 (R ² = 0.9017), respectively, and the DL was 0.5 pg/mL.

4.

(A) Schematic diagram of the configuration of ECL imaging. (B) ECL images for detecting varying concentrations of cTnI (B₁–B₄: 1, 10, 100, and 500 pg/mL). Scale bar: 30 μm. (C) Number of bright dots vs cTnI concentration.

In this work, three detection modes (ECL intensity, ECL imaging, and fluorescence imaging) were employed to quantify the target cTnI. By comparing with traditional ECL intensity methods (1.2 pg/mL in this work), fluorescence (0.001 ng/mL), ECL (4.5 pg/mL), transmitted surface plasmon resonance immunoassay (0.079 ng/mL), and CL immunoassay (1.57 pg/mL), the digital imaging method offers a lower detection limit and a wider linear range in this work (Table S3). The good analytical performance is ascribed to the quantification mechanism, in which traditional ECL intensity is based on “analog” approaches where the average signal from an ensemble of labeled protein molecules is measured, whereas the digital imaging method is based on single-molecule reaction events by counting single-particle labels bound to immunocomplexes. Compared with the two digital imaging methods, the fluorescence imaging method displays reasonable sensitivity and a lower detection limit for cTnI analysis (Table S3). The limited ECL imaging method in this work should be ascribed to the limited beads analyzed ratio because only one image is analyzed due to the oxidation of the thiol group up to +0.8 V potential. Despite the limited beads analyzed ratio, the detection limit is nearly 2 orders of magnitude lower than the cutoff value of cTnI (70–400 pg/mL) for the diagnosis of AMI. Different linear ranges and detection limits in this work may be promising for different application scenarios.

Real Sample Analysis

To demonstrate the possibility of the analytical application of the developed TDF-bridged aptasensing method, six authentic human serum samples were collected and analyzed using our imaging method. The results are summarized in Table . There is no significant difference in the average cTnI levels between the proposed FL imaging method and the clinical chemiluminescence immunoassay method (t-testing; Figure S26A). Although there is a discrepancy between our method and the CL immunoassay for two samples (samples 2 and 3), a good correlation (the correlation coefficient attained 0.9822) was obtained between the data from this method and those from the chemiluminescence immunoassay (Figure S26B). Additionally, the recovery test was conducted for two samples (<50 pg/mL, samples 4 and 6). The recoveries were calculated to be in the range of 95–114% (Table S4), which is promising for the real sample assay at lower levels of cTnI.

1. Analytical Results of cTnI in Human Serum Samples.

Sample	Sex	Age	Reason for Referral	Concentration Detected by This Method (pg/mL)	Concentration Detected by Chemiluminescence Immunoassay (pg/mL)
1	Female	29	Health Checkup	277.3	230
2	Female	59	Hyperlipidemia	68.4	110
3	Male	73	Postgastrectomy (for tumor)	8.2	70
4	Male	73	Elevated Blood Glucose	6.3	<50
5	Male	72	Hypertension	167.5	140
6	Male	34	Health Checkup	8.2	<50

Conclusions

We developed a TDF-bridged aptasensing method for the ultrasensitive determination of cTnI via ECL and fluorescence imaging methods. A specific anti-cTnI aptamer-tagged TDF was designed as a capture probe by incorporating a specific aptamer against cTnI within a three-dimensional tetrahedral DNA framework and was modified onto a gold electrode surface to provide a specific platform for cTnI with a 3D binding space at the nanoscale. By further using a nanomaterial-based microsized signal probe (Ab2-biotin•SA@MB-Ru1) and a sandwich assay, cTnI can be determined with desirable selectivity and high sensitivity using ECL intensity detection in the range of 5 to 80 pg/mL. Furthermore, we proposed a dual FL and ECL imaging method based on the TDF-bridged aptasensing platform for the ultrasensitive determination of cTnI. The results displayed that the particle number in FL and ECL images showed an increasing trend with the rising concentration of cTnI from 0.5 to 500 pg/mL. Further work can focus on the selection of signal carriers without autofluorescence, the improvement of bead analysis ratios in the digital imaging method, and the optimization of experimental conditions. This work demonstrates that the combination of a TDF-bridged aptasensing interface with the imaging method will provide a promising approach for the sensitive assay of biomarkers.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/cbmi.5c00292.

• Reagents and apparatus, preparation of the signal probe, imaging analysis with ImageJ, FL, AFM, CLSM, FL, and ECL images, analytical results (PDF)

#.

J.C. and W.Z. contributed equally to this work.

Ethics declarations: The human serum samples were provided by Shaanxi Normal University Hospital. This work was approved by the Ethics Committee of Shaanxi Normal University (No. 202610027).

The authors declare no competing financial interest.