Abstract
Osteosarcoma (OS) is the most common primary malignant bone tumor in adolescents. Diagnosis usually comes late because micrometastases are hard to see on imaging or in biopsy specimens. Exosomal microRNAs survive well in body fluids and have therefore become popular markers for liquid biopsy. miR-181b is abnormally expressed in OS, yet it occurs at low levels in body fluids and differs from miR-181a, miR-181c, and miR-181d by only a few bases. Measuring it demands both high sensitivity and high specificity. Here, we describe a label-free electrocatalytic electrochemical biosensor that pairs a peptide nucleic acid (PNA) capture probe with the exponential amplification reaction (EXPAR). A 3′-thiolated PNA probe was attached to a gold electrode by rapid freezing-assisted Au–S bonding. EXPAR converted the target into single-stranded DNA, which bound PNA and a signal probe to form a sandwich. PNA carries a neutral backbone, whereas DNA is negatively charged, so [Ru(NH3)6]3+ was adsorbed onto the sandwich by electrostatic attraction. [Fe(CN)6]3− then re-oxidized the reduced Ru complex in a catalytic cycle, and square wave voltammetry recorded the current. Under optimized conditions, the sensor responded linearly to miR-181b from 1 fM to 1 nM, with an experimental detection limit of 1 fM. It separated miR-181b from miR-181a, miR-181c, and miR-181d, and the inter-day RSD over 3 consecutive days was 4.34% (n = 9; three electrodes per day). After 6 days of storage in PBS at 4 °C, the electrodes still delivered 83.8% of their initial response. All measurements used synthetic miR-181b. Clinical validation and long-term shelf-life stability have not yet been established.
1. Introduction
Osteosarcoma (OS) is the most common primary malignant bone tumor, and it affects mainly adolescents and older adults. Neoadjuvant chemotherapy, surgery, and adjuvant chemotherapy together have raised the 5-year survival rate to about 60–70%. For patients who already have metastases at diagnosis, however, survival is only 20–30% [1,2,3]. Metastasis remains the leading cause of death, and early-stage disease often produces no specific symptoms, so micrometastases escape both imaging and pathological biopsy [4]. Tests that are non-invasive, sensitive, and specific would help with early diagnosis and with monitoring of metastasis.
Liquid biopsy looks for tumor-derived markers in body fluids and offers one route to earlier diagnosis. Exosomes are extracellular vesicles of about 30–150 nm, bounded by a lipid bilayer and present in blood and other body fluids [5,6,7,8]. They carry proteins and nucleic acids from the parent cell, and the bilayer shields miRNA from RNase digestion. Exosomal miRNA is thus more abundant and more stable than free plasma miRNA, which makes it an attractive marker for cancer liquid biopsy [8,9]. miR-181b is abnormally expressed in OS and has been linked to tumor progression and metastasis, so it is a candidate diagnostic marker [10,11]. Two properties make it hard to measure: it occurs at very low levels in body fluids, and its sequence is close to that of miR-181a, miR-181c, and miR-181d. Any assay for miR-181b must therefore be both sensitive and specific.
Northern blot, microarray, and qRT-PCR remain the standard methods for miRNA detection, but they offer low throughput, high cost, complicated primer design, or false-positive results [12,13]. Isothermal amplification has pushed sensitivity much higher, either with enzymes [14,15], without enzymes [16,17], or with nanomaterials [18,19]. Enzyme-assisted formats include rolling circle amplification (RCA) [20,21], strand displacement amplification (SDA) [22,23], loop-mediated isothermal amplification (LAMP) [24,25], and the exponential amplification reaction (EXPAR) [26,27]. EXPAR is simple to run and yields large amounts of product within a short time [28]. In EXPAR, the linearly amplified oligonucleotide (ssDNA) creates new primer–template duplexes and thereby starts a cascade. Products still need an efficient and specific way of generating a signal, however, if ultrasensitive detection in complex samples is the goal.
Electrochemical biosensors are low-cost, fast, and easy to miniaturize. Label-free formats skip labelling altogether and follow hybridization through a redox indicator [29,30]. [Ru(NH3)6]3+ binds by electrostatic attraction to a negatively charged DNA duplex, and adding [Fe(CN)6]3−/[Fe(CN)6]4− closes an electrocatalytic cycle that amplifies the signal substantially [31,32]. DNA capture probes are themselves negatively charged, however, so non-specific electrostatic adsorption is hard to avoid and the background rises.
Peptide nucleic acid (PNA) is a synthetic analogue of nucleic acid, built on a neutral N-(2-aminoethyl)glycine backbone instead of the sugar–phosphate backbone of DNA [33]. PNA:DNA and PNA:RNA duplexes are more stable than DNA:RNA and RNA:RNA duplexes, and their stability changes little with ionic strength, solvent polarity, or chaotropic agents. The non-natural backbone also resists nucleases and proteases, so PNA probes keep longer than natural oligonucleotides [34,35,36]. These properties make PNA a good capture probe for electrochemical sensing. Pairing a PNA capture layer with EXPAR and with Ru/Fe electrocatalytic amplification should give a sensitive and specific platform for miRNA detection.
In this work, we built a label-free electrocatalytic sensor based on PNA and used it to detect miR-181b, a potential biomarker for osteosarcoma. Synthetic miR-181b served as the model target. PNA was immobilized on a gold electrode through its 3′-thiol group, and EXPAR converted the target into single-stranded DNA. Most reported electrochemical miRNA assays depend on covalently labelled probes or on enzyme-tethered tags, and both add cost and preparation steps. Here, the capture layer is neutral PNA, which suppresses the electrostatic background that a DNA capture layer produces. The readout is label-free as well: [Ru(NH3)6]3+ is held electrostatically on the PNA:DNA:SP duplex, and a Ru/Fe cycle recycles the Ru(II) centre and amplifies the current without any labelling step. The amplification products bound PNA and a signal probe (SP) to form a sandwich, which adsorbed [Ru(NH3)6]3+ and was then amplified through [Fe(CN)6]3−/[Fe(CN)6]4− cycling and measured by square wave voltammetry (SWV). The sensor was linear from 1 fM to 1 nM and separated miR-181b from miR-181a, miR-181c, and miR-181d.
2. Materials and Methods
2.1. Reagents and Apparatus
NaCl, Na2HPO4·12H2O, NaH2PO4·2H2O, HCl, K3[Fe(CN)6], and Tween-20 were obtained from Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Tris-HCl, TE powder, and TBE buffer were purchased from Sangon Biotech (Shanghai, China). Anhydrous ethanol was obtained from Xilong Scientific Co., Ltd. (Shantou, China), Al2O3 powder from Buehler (Lake Bluff, IL, USA), and concentrated H2SO4 from Shanghai Sanying Chemical Reagent Co., Ltd. (Shanghai, China). [Ru(NH3)6]Cl3 was purchased from Merck (Darmstadt, Germany).
Electrochemical measurements were performed on a CHI1040c workstation (Shanghai Chenhua Instrument Co., Ltd., Shanghai, China) with a three-electrode system consisting of an AuE working electrode (radius 2 mm), a platinum wire counter electrode, and an Ag/AgCl (3 M KCl) reference electrode (Shanghai Chenhua Instrument Co.,Ltd., Shanghai, China). Other apparatus included a Direct-Q water system (Millipore, Burlington, MA, USA), a thermal cycler (Hangzhou Bioer Technology Co., Ltd., Hangzhou, China), and an electrophoresis system (Bio-Rad Laboratories, Hercules, CA, USA).
2.2. Oligonucleotide Sequences
Sequences of PNA, SP, miR-181b, miR-181a, miR-181c, miR-181d, DNA target, and EXPAR template are listed in Table 1.
Table 1.
Sequence of oligonucleotides used in this work.
2.3. EXPAR Amplification
The EXPAR reaction was performed in a total volume of 30 μL, prepared on ice in the indicated order. Before use, the EXPAR-template was annealed by adding an appropriate amount of Tris-HCl as annealing buffer, heating at 95 °C for 5 min, and slowly cooling to room temperature. The reaction system contained 10 mM of Tris-HCl buffer, 1× ThermoPol Reaction Buffer, 0.9× NEBuffer r3.1, 0.3 U/μL Nt.BstNBI, 0.03 U/μL Vent(exo-) DNA polymerase, 500 μM of dNTPs, 100 nM of EXPAR template, and different concentrations of target. The reaction mixture was incubated at 55 °C for 1 h in a thermal cycler. After amplification, the products were heat-inactivated at 80 °C for 20 min to denature Vent(exo-) and Nt.BstNBI, then diluted in PBS and used directly without further purification.
2.4. Electrode Pretreatment and PNA Immobilization
AuE was polished to a mirror-like surface with 0.05 μm Al2O3 suspension for 3 min, rinsed with distilled water, ultrasonicated in anhydrous ethanol and distilled water for 30 s each, and then scanned in 0.5 M H2SO4 from 0 to 1.6 V until the electrochemical response was stable. After rinsing, the electrode was dried under nitrogen. PNA probe modified with a 3′-thiol was diluted to 2 μM in TE containing TCEP and dropped onto the AuE surface. PNA was attached to the electrode via Au–S bonding by freezing the droplet, while evaporation during immobilization was prevented. When the droplet is frozen, ice nucleation excludes the PNA and dissolved salts from the growing ice lattice, concentrating both in the unfrozen liquid channels that remain in contact with the gold surface. The local PNA concentration therefore increases by orders of magnitude, which accelerates the Au–S chemisorption kinetics and favors a dense, uniformly packed monolayer within 1 h. The same effect prevents evaporation of the small droplet and limits thiol oxidation during immobilization; this approach is known to accelerate thiol adsorption on gold [37]. We compared rapid freezing at −20 °C for 1 h with overnight incubation (12 h) at room temperature. Freezing gave a denser PNA layer, judged from the higher charge-transfer resistance of the resulting PNA/AuE, and it cut electrode preparation from 12 h to 1 h. Blocking followed with 2 mM MCH in the dark for 1 h.
2.5. Sandwich Structure Formation
The blocked electrode was washed with 10 mM of PB and incubated in 100 μL of PBS (10 mM PB, 0.1 M NaCl, pH 7.4) containing EXPAR products and 200 nM of SP at 45 °C for 1 h. The control group was incubated under identical conditions with 200 nM of SP in the same volume of PBS containing the EXPAR reaction mixture without amplification products (containing only the amplification template, heat-inactivated enzyme, dNTPs, and other components carried over from the previous reaction) to allow SP hybridization.
2.6. Electrochemical Measurements and Data Analysis
The detection electrolyte contained 500 nM of and 750 µM of K3[Fe(CN)6] in 10 mM of Tris-HCl (pH 7.4) containing 7.5 mM of NaCl (total ionic strength ≈ 20 mM). Electrodes were immersed for 5 min before measurement. SWV ran from +0.4 to −0.8 V, with a potential increment of 4 mV, an amplitude of 25 mV, and a frequency of 25 Hz. The SWV baseline was defined by linear (tangent) extrapolation of the current over a faradaic-free potential window (+0.05 to +0.35 V for the [Ru(NH3)6]3+ reduction peak; −0.45 to −0.75 V for the [Fe(CN)6]3− reduction peak), and the peak current was measured as the height of the peak maximum above this baseline. The same windows were used for every concentration and every electrode, so capacitive and diffusion background contributions are subtracted consistently across the calibration series. Amperometric i–t curves were recorded at −0.1 V for 100 s. Electrochemical impedance spectroscopy (EIS) data were analyzed using ZView software (Version 3.1, Scribner Associates). The Nyquist plots were fitted to an appropriate equivalent circuit using the complex nonlinear least-squares (CNLS) method. The goodness of fit was evaluated by the chi-square (χ2) parameter, which was on the order of 10−4 Additionally, the standard errors for the fitted elements (such as Rs and Rct) were monitored to ensure they remained within acceptable limits (relative errors < 5%). All measurements were made on a CHI1040c workstation with a three-electrode system, and data are given as mean ± standard deviation (n = 3 unless stated otherwise). Five blank electrodes (MCH/PNA/AuE without target) gave 292.0 ± 46.1 nA (n = 5) under the same SWV conditions, so a signal was scored positive only above 430.3 nA (blank mean + 3 SD). The current at 1 fM was above this value, and no lower concentration was tested, so 1 fM is reported as the detection limit. The 1 fM current (n = 3) was also compared with the blank (n = 5) by a two-tailed Welch’s t-test, which does not assume equal variances. The difference was highly significant (p < 0.001), so the 1 fM response is statistically distinct from the blank.
3. Results
3.1. Construction Principle and Signal Transduction Mechanism of the PNA-Based Label-Free Electrocatalytic Electrochemical Sensor
Figure 1 outlines the design. A PNA probe complementary to miR-181b carries a thiol group (–SH) at the 3′ end, and this thiol binds covalently to bare gold through an Au–S bond, which anchors the PNA on the electrode. MCH then blocked the remaining gold surface, limiting non-specific adsorption and improving sensor stability.
The EXPAR template has three regions: a loop-recognition region, a stem-pairing region, and a non-pairing arm that carries trigger strand X and the recognition site for the nicking endonuclease. Without target, stem pairing keeps the template closed. Target binding opens it in four steps. (1) The target binds the exposed toehold and opens the hairpin by toehold-mediated strand displacement. (2) DNA polymerase extends the 3′ end of the bound target along the template to form a fully complementary duplex. (3) Duplex formation restores a complete Nt.BstNBI recognition site, and nicking of the newly synthesized strand releases trigger strand X together with the extended target. (4) Released X binds a fresh template and starts another round of extension and nicking. Cycling of X and of the displaced target product makes the amplification exponential.
PNA on the electrode then hybridizes with the ssDNA amplification product, and SP binds another region of the same strand, forming a sandwich. Because PNA is electrically neutral while DNA is negatively charged, the duplex adsorbs a large amount of Ru(NH3)63+ from Tris-HCl solution, and reduction of the bound cation gives peaks that CV and SWV can record. In the designed alignment, PNA covers nt 1–13 of miR-181b and SP covers the remaining nt 14–23, so the two probes sit side by side and jointly span the full 23 nt amplification product; the sequences in Table 1 define this alignment. Adding K3[Fe(CN)6] then allows Ru(NH3)62+ + [Fe(CN)6]3− ⇌ Ru(NH3)63+ + [Fe(CN)6]4−, which re-oxidizes the reduced Ru complex and amplifies the signal.
Figure 1.
Schematic illustration of the PNA−based label−free electrocatalytic electrochemical sensor. The dashed box shows the EXPAR hairpin model and the amplification cycles. The Roman numerals mark the reaction sequence: (I) target binding to the toehold and hairpin opening by strand displacement; (II) DNA−polymerase extension of the target to form a complete duplex; (III) restoration of the Nt.BstNBI site and nicking to release trigger X and the extended target; (IV) binding of released X to a fresh template; (V−VII) repeated extension and nicking in the trigger X cycle; and (VIII) recycling of the displaced target product, giving exponential amplification. The lower panel shows MCH/PNA assembly on the gold electrode (GE), target−dependent sandwich formation, Ru(NH3)63+ adsorption, and the Ru/Fe electrocatalytic SWV readout with the Ru and Fe peaks.
3.2. Feasibility Validation of Sandwich Assembly, Electrostatic Adsorption, and Ru/Fe Electrocatalytic Signal Amplification
3.2.1. Stepwise Electrochemical Characterization of the Electrode Assembly by EIS and CV
Each assembly step was followed by EIS and CV in 5 mM [Fe(CN)6]3−/4− containing 0.1 M KCl (CV at 50 mV/s; EIS from 0.1 Hz to 100 kHz with a 5 mV amplitude). Six configurations were compared: (a) bare AuE; (b) PNA/AuE prepared at room temperature; (c) PNA/AuE prepared by rapid freezing at −20 °C for 1 h; (d) MCH/PNA/AuE; (e) MCH/PNA/AuE after capture of the EXPAR product; and (f) the complete PNA:amplicon:SP sandwich after hybridization with SP. Nyquist plots in Figure 2A changed in clear steps. PNA immobilization raised the charge-transfer resistance (Rct) above that of bare AuE (curves a–c), which shows successful anchoring of the thiolated PNA through Au–S bonds. Freezing (c) produced a larger Rct than room-temperature incubation (b), so the frozen protocol gave denser and more complete PNA coverage, and it did so in 1 h rather than 12 h. MCH backfilling raised Rct again (d), as expected for a monolayer that blocks the free gold surface. Capturing the EXPAR product raised it further (e), and hybridization with SP gave the highest value of the series (f). This last increase shows that the extra negative charge of the PNA:amplicon:SP duplex hinders access of [Fe(CN)6]3−/4− to the electrode. CV curves in Figure 2B followed the same trend: the peak-to-peak separation ΔEp widened step by step and the peak current fell from curve a to curve f. Since [Fe(CN)6]3−/4− is anionic, any step that adds negative charge or blocks gold registers as an increase in Rct (a → b/c → d → e → f), so PNA anchoring, MCH backfilling, amplicon capture, and SP hybridization each proceeded as designed.
Figure 2.
Stepwise characterization of the electrode assembly and of the individual contributions of Ru(NH3)63+ and [Fe(CN)6]3−. (A) Nyquist plots recorded in 5 mM of [Fe(CN)6]3−/4−containing 0.1 M of KCl (0.1 Hz-100 kHz, 5 mV amplitude) for (a)−(f); the inset shows the Randles equivalent circuit used for fitting (Rs in series with a parallel combination of the constant phase element CPE and (Rct + Wo)) and (B) CV curves recorded in 5 mM of [Fe(CN)6]3−/4− containing 0.1 M of KCl (CV at 50 mV/s; EIS from 0.1 Hz to 100 kHz, 5 mV amplitude) for (a) bare AuE; (b) PNA/AuE prepared by room−temperature incubation; (c) PNA/AuE prepared by rapid freezing−assisted immobilization at −20 °C for 1 h; (d) MCH/PNA/AuE; (e) MCH/PNA/AuE after capture of the EXPAR product; and (f) the complete PNA:amplicon:SP sandwich obtained by further hybridization with SP. (C) SWV curves of the PNA/MCH−modified AuE in Tris−HCl containing 100, 500, and 1000 nM Ru(NH3)6Cl3, without (blank) and with the DNA target. (D) Corresponding peak currents and signal−to−background (S/B) ratios. (E) SWV curves in 500 nM Ru(NH3)6Cl3/750 μM K3[Fe(CN)6] without and with the DNA target. (F) SWV curves in the same electrolyte without and with the target miR−181b (RNA).
The fitted values behind Figure 2A are collected in Table 2, and the corresponding CV parameters are in Table 3. Rct rose from 75.08 Ω for bare AuE to 410 Ω for (b), 535.9 Ω for (c), 1705 Ω for (d), 2163 Ω for (e), and 4415 Ω for (f), a 59-fold increase in total. Rs stayed between 38.5 and 53.1 Ω (46.8 ± 5.4 Ω) and showed no trend along the series, so the changes are interfacial and not caused by the electrolyte. Freezing raised Rct by 31% over room-temperature incubation. If the residual electroactive area is taken as Rct(a)/Rct(mod), it fell from 18.3% for (b) to 14.0% for (c). CV gave the same order: the anodic peak current fell from 31.26 μA (a) to 23.16 μA (f), and ΔEp widened from 70 mV to 172 mV (Table 3).
CV separates (b) and (c) less clearly than EIS (30.60 versus 29.80 μA, 2.6%). At 50 mV/s, the diffusion layer is about 35 μm thick, four orders of magnitude more than the PNA layer, so the peak current reports the area that the probe can reach rather than the fraction of gold that is blocked. Rct is measured at the interface and resolves the two protocols. We therefore use Rct as a relative measure of blocking. The absolute surface density of the PNA probe was not determined, and this is stated as a limitation in Section 3.7.
Table 2.
Fitted parameters of the equivalent circuit used for the Nyquist plots of configurations (a)–(f) in 5 mM [Fe(CN)6]3–/4– containing 0.1 M KCl. Nyquist plots (0.1 Hz–100 kHz, 5 mV amplitude) were fitted to Rs(CPE∥(Rct + Wo)) (inset of Figure 2A) with ZView 3.1 by the CNLS method. CPE–T is the prefactor Q in S sn and CPE-P its exponent n.
Table 3.
Cyclic voltammetric parameters of the six electrode configurations (a)–(f) in 5 mM [Fe(CN)6]3−/4− containing 0.1 M KCl at 50 mV/s. Each value is the mean of duplicate scans; ΔEp = Epa − Epc.
3.2.2. Feasibility of Electrostatic Adsorption and Ru/Fe Electrocatalytic Signal Amplification
To test the sensing strategy, we designed a PNA probe with a 3′-terminal thiol as the capture probe (CP) for miR-181b. PNA/MCH-modified AuE was hybridized with a DNA strand of the same base sequence as miR-181b (U replaced by T) and then placed in Tris-HCl containing Ru(NH3)6Cl3 for SWV. Figure 2C compares the SWV responses at 100, 500, and 1000 nM Ru(NH3)6Cl3, and Figure 2D summarizes peak currents and signal-to-background (S/B) ratios. With the DNA target present, PNA and SP formed a negatively charged PNA:DNA:SP sandwich that adsorbed Ru(NH3)63+ and gave a Ru reduction peak much higher than that of the blank. Without target, the signal stayed low and came mainly from free Ru complex diffusing to the gold surface. Currents rose with Ru concentration, but so did the diffusion background. The highest S/B ratio appeared at 500 nM Ru(NH3)6Cl3 (Figure 2D), where adsorption on the duplex was nearly saturated while the diffusion background was still low, and we used this concentration in all later measurements.
In addition to the redox probe concentration, the ionic strength of the detection buffer is another crucial factor for electrostatic adsorption. In our optimized protocol, the final electrolyte has a relatively low total ionic strength of approximately 20 mM (10 mM Tris-HCl, pH 7.4, plus 7.5 mM NaCl introduced from the K3[Fe(CN)6] stock solution). This low-ionic-strength environment was deliberately selected. If the background salt concentration were too high, abundant cations (e.g., Na+, K+) would cause severe electrostatic screening and competitively bind to the negatively charged PNA:DNA:SP duplex. This competition would drastically hinder the electrostatic adsorption of , thereby suppressing the signal. Conversely, the trace amount of NaCl (7.5 mM) ensures the stability of the nucleic acid duplex without compromising the Ru-DNA electrostatic interaction, thus maximizing the signal-to-noise ratio.
K3[Fe(CN)6] was then added to the Ru-containing electrolyte to build the Ru/Fe electrocatalytic cycle. Without a target, no sandwich formed and SWV gave a low Ru peak and a high Fe peak. With the DNA target present, the PNA:DNA:SP sandwich adsorbed Ru(NH3)63+, and SWV gave a high Ru peak and a low Fe peak (Figure 2E). When the authentic target miR-181b (RNA) was tested, it produced an analogous reciprocal electrocatalytic response (Figure 2F), confirming that the RNA target undergoes the EXPAR process and assembles into the sandwich structure despite potential differences in absolute current magnitudes due to distinct enzyme extension kinetics on RNA versus DNA primers. The electrocatalytic cycle accounts for this. Ru(NH3)63+ bound to the duplex is reduced to Ru(NH3)62+ at the electrode and is then rapidly re-oxidized to Ru(NH3)63+ by [Fe(CN)6]3− in solution. Ru therefore re-enters the electrode reaction many times, which amplifies the reduction current, while [Fe(CN)6]3− is converted to [Fe(CN)6]4− and its own reduction peak falls. Such a reciprocal change is direct evidence for Fe(III)-mediated re-oxidation of Ru(II) and confirms that the pair works as an electrocatalytic amplifier.
3.2.3. Individual Roles of [Ru(NH3)6]3+ and [Fe(CN)6]3− in Signal Generation
The two redox species do different jobs and cannot replace each other. [Ru(NH3)6]3+ is a cation that binds the negatively charged PNA:DNA:SP duplex by electrostatic attraction, so it reports target binding directly. In 500 nM of [Ru(NH3)6]Cl3 alone, the sandwich electrode gave a Ru peak far above that of the MCH-blocked electrode (Figure 2C,D), and once the surface was saturated, the current tracked the amount of surface-bound duplex rather than the Ru concentration in solution. Each adsorbed Ru centre is reduced only once per scan, so the response is stoichiometric and limited by the number of cationic sites the duplex can accommodate. Ru(III) thus turns a hybridization event into a current, and the contrast between neutral PNA and negatively charged DNA is what makes the assay specific at the level of signal generation.
[Fe(CN)6]3− behaves differently. It is anionic, it stays in solution, it does not bind the duplex, and it senses only how permeable the organic layer is to a negative ion, which is why it served as the EIS/CV probe in Section 3.2.1. Without Ru, target binding lowers the Fe signal instead of raising it because the sandwich adds negative charge and steric hindrance. Fe peak current fell stepwise from the MCH/PNA electrode through amplicon capture to the complete sandwich (Figure 2B, curves d–f), in parallel with the impedance increase over the same steps (Figure 2A). Fe alone therefore reports the state of the surface, but with the wrong sign for a signal-on assay, and it gives no amplification.
Amplification needs both species, and its signature is the reciprocity of the two responses. In the Ru/Fe electrolyte, the Ru peak of the sandwich electrode rose about six-fold above the blank, while the Fe peak fell to about 15%, and the two changes were anti-correlated across blank, DNA target, and miR-181b target (Figure 2E,F). Such reciprocity is typical of an electrocatalytic (EC′) mechanism. Ru(NH3)63+ confined on the duplex is reduced at the electrode and then re-oxidized chemically by [Fe(CN)6]3− in solution rather than at the electrode. The same Ru centre therefore turns over many times during a single scan, while Fe(III) is consumed and its own reduction peak must fall. Ru supplies both the charge that reports recognition and the catalytic centre; Fe acts as the sacrificial oxidant; neither alone can amplify.
3.3. Feasibility Validation of the EXPAR Amplification System
EXPAR was checked by gel electrophoresis and by i–t measurements. In Figure 3, lane 1 was the 25–500 bp DNA marker, lane 2 the blank, lane 3 the reaction without nicking enzyme, lane 4 the reaction without DNA polymerase, lane 5 the reaction without EXPAR hairpin template, and lane 6 the target group. Product ran between 25 and 50 bp, as expected from its size. Leaving out any one component stopped the cycle, so all components are required. Gel electrophoresis shows only that the target group produces a product of the expected size; it does not show whether that product can be captured by PNA and hybridized with SP at the same time. Running Ru/Fe SWV at this stage would not settle the question either. The signal would come from Ru3+ adsorbed on the negatively charged duplex, and the EXPAR mixture also contains DNA polymerase, nicking enzyme, template, dNTPs, and abundant non-target nucleic acid. All of these may adsorb on the electrode or bind Ru3+, so an SWV signal could not be assigned unambiguously to the PNA−amplicon−SP sandwich. To avoid this ambiguity, we used a 5′−biotinylated SP with SA-HRP−catalysed TMB chronoamperometry (i−t) as an orthogonal readout. HRP attaches to the electrode only when the EXPAR product is captured by PNA and hybridized with biotin−SP, and the TMB reduction current does not depend on Ru3+ adsorption or on the charge of the amplicon. This gives a more specific check that SP is part of the sandwich. Figure 3B shows a clear current difference between the target group (EXPAR product containing miR−181b) and the blank group (EXPAR product without target), so the product can indeed form the sandwich with CP and SP. i−t served only as an auxiliary check here; all quantitative detection used Ru/Fe electrocatalytic SWV.
Figure 3.
Feasibility analysis of the EXPAR amplification system. (A) Gel electrophoresis image: The amplification product is indicated by an arrow. (B) SA−HRP/TMB chronoamperometric (i−t) control confirming that the EXPAR product can form the PNA−ssDNA−biotin−SP sandwich.
3.4. Systematic Optimization of Key Experimental Conditions for the Sensor
Detection conditions were optimized one at a time: PNA concentration, hybridization temperature and time, EXPAR template concentration, amplification time, nicking enzyme concentration, DNA polymerase concentration, and the Ru/Fe ratio of the detection solution. PNA concentration was optimized with the auxiliary SA-HRP/TMB i–t readout of Section 3.3. HRP/TMB steady-state current reports how much sandwich can capture the enzyme, and as an enzymatic signal it does not depend on electrostatic adsorption, so it ranks PNA loading stably and with low noise. All other conditions, and all analytical figures of merit, were obtained by SWV, the readout used in the final assay.
i–t signal was best at 2 μM of PNA (Figure 4A). Higher concentrations packed the probes too densely, which hindered access of target ssDNA and SP to PNA and lowered the hybridization efficiency. Below 1 μM, there were too few capture probes, fewer sandwiches formed, and the signal dropped, so we used 2 μM. Signal peaked at 45 °C (Figure 4B). Higher temperatures destabilized the sandwich, and lower ones reduced the specificity of base pairing; 45 °C is close to the melting temperature of PNA:DNA duplexes, so it gave specific hybridization without dissociating the duplex. Signal levelled off at 60 min (Figure 4C), while the background rose slowly with longer incubation, which means hybridization was complete at 60 min and longer times only added non-specific adsorption. Hybridization therefore ran for 60 min.
Figure 4.
Optimization I. (A) PNA concentration, read out by the auxiliary SA−HRP/TMB chronoamperometric (i−t) steady−state current; (B) hybridization temperature; (C) hybridization time; (D) EXPAR template concentration, with (B–D) read out by the SWV peak current in the Ru/Fe electrolyte. Error bars, n = 3.
Target-to-background difference was largest at 100 nM EXPAR template (Figure 4D). At 125 nM, the nicking enzyme could not cleave the duplex completely, less single-stranded product formed, and the signal fell; among 50, 75, 100, and 125 nM, 100 nM was best. Signal peaked at 60 min of amplification and decreased at 80 min with a higher background (Figure 5A), probably because too much product was bound to the template and non-specific amplification increased. Nicking enzyme worked best at 0.3 U/μL (Figure 5B); lower concentrations cleaved too little and higher ones increased non-specific cleavage. DNA polymerase at 0.03 U/μL gave a lower background (Figure 5C), whereas higher concentrations caused non-specific amplification, so we kept 0.3 U/μL nicking enzyme with 0.03 U/μL polymerase. Ru/Fe ratio was optimal at 1:1500 (Figure 5D). At lower Fe3+, the re-oxidation of Ru2+ was too slow and amplification was limited; at higher Fe3+, the Fe reduction current itself rose, which raised the background and suppressed the Ru response. Of 1:500, 1:1000, 1:1500, 1:2000, and 1:2500, we chose 1:1500.
Figure 5.
Optimization II. (A) EXPAR reaction time; (B) nicking enzyme concentration; (C) DNA polymerase concentration; (D) Ru(NH3)63+/[Fe(CN)6]3− ratio. Error bars, n = 3.
3.5. Analytical Performance of the Sensor
3.5.1. Sensitivity
Under these conditions, SWV was used to measure different concentrations of synthetic miR-181b, and the current rose with target concentration (Figure 6A). From 1 fM to 1 nM, current I (nA) was linear against the logarithm of miR-181b concentration: I = 176.54 lgCtarget (aM) + 337.05 (R2 = 0.9813; five concentration levels, duplicate measurements; Figure 6B). The current measured at 1 fM differed significantly from the blank (p < 0.001, Welch’s t-test; blank mean + 3SD = 430.3 nA, Section 2.6), and no lower concentration was tested, so 1 fM is reported as the detection limit. The calibration curve was not extrapolated below this point. A linear range of 1 fM–1 nM covers six orders of magnitude and overlaps the reported concentration range of circulating miRNA in serum, which suggests that the assay could be applied to real samples.
Figure 6.
Analytical performance of the sensor. (A) SWV curves at different concentrations of synthetic miR−181b (1 fM−1 nM). (B) Calibration plot of peak current versus the logarithm of the synthetic miR−181b concentration. (C) SWV comparison for synthetic miR−181b, miR−181a, miR−181c and miR−181d. (D) Specificity bar chart. (E) SWV responses of nine individual electrodes, three measured on each of 3 consecutive days, for the EXPAR products generated from 1 nM of synthetic miR−181b. (F) Reproducibility bar chart showing the peak currents of the nine electrodes, three per day, and their mean. (G) SWV responses recorded on the preparation day (day 0) and after 3 and 6 days of storage in PBS at 4 °C, measured with EXPAR products generated from 1 nM of synthetic miR−181b (three electrodes per time point). (H) Mean peak currents (bars, left axis) and signal change (%, line, right axis) over the storage period; the response retained 83.8% of the day 0 value after 6 days. Error bars represent the standard deviation (n = 9).
3.5.2. Specificity
Members of the miR-181 family have very similar sequences, so specificity matters here. Under the optimized conditions, 1 nM of miR-181a, miR-181c, and miR-181d was measured as interferents and compared with an equimolar concentration of miR-181b. Signal for miR-181b was much higher than for the other three (Figure 6C,D), and in Figure 6D, the peak currents of the three interferents all fell in a narrow band well below the miR-181b response. Residual current of miR-181d (about 570 nA) was about 30% of the equimolar miR-181b current (about 2000 nA), and about 18% after blank subtraction (blank, about 250 nA); miR-181a and miR-181c gave lower responses still. This cross-reactivity is a limitation, because in samples where miR-181d is present in large excess, it could produce a false positive. A sequence was scored positive when its signal exceeded blank + 3SD and reached at least 70% of the equimolar miR-181b response, and by this criterion, all three interferents scored negative, including miR-181d, which differs from miR-181b at a single position (C → U at position 12). Selectivity comes from two sources. The EXPAR template is sensitive to mismatches in the region that primes extension, so a single-base difference sharply lowers amplification efficiency and the amount of product available for sandwich formation. PNA hybridization then adds stringency: a single mismatch lowers the melting temperature of the short PNA:RNA duplex and destabilizes the sandwich at 45 °C, while the neutral backbone limits the non-specific electrostatic adsorption that would otherwise raise the blank. Where a homologous isoform predominates and higher stringency is needed, hybridization temperature can be raised within 45–55 °C and ionic strength adjusted accordingly.
3.5.3. Reproducibility
Reproducibility was tested on 3 consecutive days. On each day, three freshly prepared electrodes measured the EXPAR products from 1 nM of synthetic miR-181b, so nine electrodes were compared in total (Figure 6E,F). Their peak currents gave a relative standard deviation of 4.34% (n = 9). Because the nine electrodes were measured on three different days, this value includes both electrode-to-electrode and day-to-day variation, which is the comparison that matters when the assay is used on different days. All electrodes came from one batch of gold electrodes, so inter-batch variability has not been examined.
3.5.4. Stability
The storage stability of the PNA-modified electrodes was examined under the conditions used for the reproducibility test, with EXPAR products generated from 1 nM of synthetic miR-181b. Electrodes were measured by SWV on the preparation day (day 0) and again after 3 and 6 days, during which they were kept in PBS at 4 °C (Figure 6G,H). The response decreased only slightly within the first 3 days (approximately 95% of the initial value) and still reached 83.8% of the day 0 signal after 6 days. The shape and peak potential of the SWV curves did not change over this period (Figure 6G). The small drop in current therefore looks like a slow loss of surface-bound PNA at the Au–S interface rather than damage to the recognition layer. Within 6 days, the electrodes kept their recognition and signal-generation capacity. A shelf-life study over weeks to months is still needed before routine use (Section 3.7).
3.6. Comparison with Reported miRNA Detection Methods
Table 4 lists representative isothermal and enzyme-assisted strategies for miRNA detection, and the present assay can be compared with them in terms of detection limit, linear range, and ease of operation. The LOD obtained here (1 fM) is one to two orders of magnitude lower than those of the CRISPR/Cas fluorescence assays listed, PER-E-CRISPR (30.2 fM) and EXPCas (103 fM), and lower than that of the APE1-driven G-quadruplex/ThT assay (25.6 fM). It is about the same as the label-free potential-assisted ternary hybridization assay (1.28 fM). Only the in situ ZIF-8 nanolabel method reaches a lower value (2.3 aM), and it does so at the cost of poly(A) tailing and nanolabel amplification, so it is no longer label-free. Linear range here (1 fM–1 nM) covers six orders of magnitude, wider than the 3 decades of PER-E-CRISPR (0.1–100 pM) and the 5 decades of the ZIF-8 nanolabel (10 aM–1 pM) and APE1/G4-ThT (10 aM–1 pM) assays. It also covers the femtomolar to nanomolar interval in which circulating miRNAs are usually measured.
Table 4.
Comparison of analytical performance for miRNA detection methods.
The clearer advantage of this work is in the readout. Three of the five methods listed rely on a label or a nanomaterial tag (an ssDNA molecular beacon, a fluorescent reporter and a ZIF-8 nanolabel), and the other two, although label-free, depend on G-quadruplex/ThT structure switching or on potential-assisted ternary hybridization. Our design is label-free throughout. Target recognition becomes current through electrostatic confinement of [Ru(NH3)6]3+ on the PNA:DNA duplex, and the Ru/Fe cycle then amplifies it. No covalent label, fluorescent reader, or signal-derivatization step is needed, and the whole procedure from amplification to readout takes about 2.2 h on a single electrode. The format therefore sits closer to a disposable, equipment-light test than the optical or nanolabel alternatives. Comparison should still be read as semi-quantitative, because the listed assays differ in signal modality. A like-for-like comparison in serum or exosomal RNA, the next step of this work, will be needed before these analytical figures can be translated into a clinical advantage.
3.7. Limitations
The study has five limitations. (i) All measurements used synthetic miR-181b, with no serum, exosomal RNA, or clinical specimen analyzed and no side-by-side comparison with RT-qPCR, so the clinical value of the assay is still unproven. (ii) The storage stability evaluated so far covers 6 days of refrigerated storage in PBS, over which the electrodes retained 83.8% of their initial response (Section 3.5.4; Figure 6G,H). A shelf-life study (dry storage at 4 °C over weeks to months) and accelerated ageing have not yet been performed, and the long-term working lifetime of the sensor, and the extent to which Au–S bond hydrolysis or PNA degradation limits it, remain unknown. This remains an important gap to close before routine or point-of-care use. (iii) Reproducibility was assessed with nine electrodes from a single batch, measured as three electrodes per day on 3 consecutive days, so day-to-day variation is included, whereas inter-batch variability has not been determined. (iv) Discrimination among the highly homologous miR-181a/c/d isoforms was shown at equimolar concentrations only; the residual cross-reactivity of miR-181d (ca. 30% of the equimolar miR-181b peak current, ca. 18% after blank correction; Figure 6D) could not be removed completely, samples with a large excess of a mismatched isoform were not tested, and stringency has not been optimized for such mixtures. Future work will cover serum, cell culture supernatant, and exosomal RNA, comparison with RT-qPCR or ddPCR, a formal storage-stability study, and a systematic evaluation of inter-batch reproducibility and cross-reactivity in mixed samples. (v) The surface density of the PNA capture layer was not measured. Blocking is therefore reported as a relative fraction from Rct rather than as an absolute coverage, and a chronocoulometric determination on the PNA:DNA duplex is left for future work.
4. Conclusions
In this study, we built a label-free electrocatalytic sensor based on PNA for the ultrasensitive detection of miR-181b. A PNA capture probe, EXPAR amplification, and [Ru(NH3)6]3+/[Fe(CN)6]3− electrocatalytic signal amplification work together here to give a label-free, simple, and sensitive route to miRNA detection. The neutral backbone and nuclease resistance of PNA allow stable attachment to gold through Au–S bonds and reliable recognition, as EIS and CV confirmed step by step. EXPAR amplifies the target exponentially at constant temperature, and the [Ru(NH3)6]3+/[Fe(CN)6]3− cycle amplifies the electrochemical signal by electrocatalysis: Ru(III) bound to the duplex is repeatedly reduced and re-oxidized by Fe(III), so the current no longer depends on the amount of adsorbed redox label. Together, the three components give a linear response to synthetic miR-181b from 1 fM to 1 nM, an experimental detection limit of 1 fM, and discrimination of miR-181b from miR-181a, miR-181c, and miR-181d. Reproducibility across 3 consecutive days was good (RSD = 4.34%, n = 9), and the electrodes retained 83.8% of their initial response after 6 days of storage in PBS at 4 °C (Figure 6G,H). This demonstrates short-term wet-storage stability only; long-term dry storage, accelerated ageing, and inter-batch reproducibility were not evaluated and are prerequisites for any point-of-care deployment. Against reported miRNA-detection methods (Table 4), the method offers a competitive experimental LOD and linear range in buffer without covalent labelling of the probe. Clinical validation and a long-term storage-stability study have not yet been done, and inter-day and inter-batch reproducibility remain to be established. Future work will focus on clinical samples, comparison with qRT-PCR or ddPCR, and a formal shelf-life study, so that the sensor can be evaluated in real biological samples. Overall, this PNA-based label-free electrocatalytic sensor gives a sensitive and specific strategy for determining miR-181b in buffer, with an experimental LOD of 1 fM and a 6-decade linear range.
Author Contributions
Conceptualization, J.C. and G.Z.; Methodology, G.C., J.Z., and Q.F.; Validation, C.Z.; Formal analysis, Q.Z.; Investigation, Q.Z.; Data curation, Y.F.; Writing—original draft, G.C.; Writing—review and editing, J.Z. and Q.F.; Visualization, G.C.; Supervision, J.C. and G.Z.; Project administration, G.Z.; Funding acquisition, G.Z. All authors have read and agreed to the published version of the manuscript.
Funding
This work was supported by the National Natural Science Foundation of China (Grant No. 82673045); the Joint Funds for the Innovation of Science and Technology, Fujian Province (Grant No. 2024Y9425); the Medical Education Integration Development Fund Project of Fujian Medical University (Grant No. 2025RH051B); the Financial Science and Technology Special Fund of Fujian Province (Grant No. BPB-2025ZGX); and the Start-up Foundation for Introducing Talents of Fujian Medical University (Grant No. XRCZX2022029).
Institutional Review Board Statement
Not applicable.
Informed Consent Statement
Not applicable.
Data Availability Statement
The data presented in this study are available from the corresponding authors upon reasonable request.
Acknowledgments
We thank Central Laboratory of the First Affiliated Hospital of Fujian Medical University for providing experimental venues, professional experimental instruments and equipments. During the preparation of this manuscript, the authors used DeepSeek-V4.1 (https://www.deepseek.com/; accessed in November 2026) to improve the grammar, clarity and readability of the text; the tool was not used in data analysis, figure generation or interpretation of results. The authors have reviewed and edited the output and take full responsibility for the content of this publication.
Conflicts of Interest
The authors declare no conflicts of interest.
Abbreviations
The following abbreviations are used in this manuscript:
| EXPAR | exponential amplification reaction |
| LOD | limit of detection |
| MCH | 6-mercapto-1-hexanol |
| NEase | nicking endonuclease |
| PB | phosphate buffer |
| PBS | phosphate-buffered saline |
| PNA | peptide nucleic acid |
| RSD | relative standard deviation |
| SWV | square wave voltammetry |
| TCEP | tris(2-carboxyethyl)phosphine |
| TE | Tris-EDTA |
References
- Beird, H.C.; Bielack, S.S.; Flanagan, A.M.; Gill, J.; Heymann, D.; Janeway, K.A.; Livingston, J.A.; Roberts, R.D.; Strauss, S.J.; Gorlick, R. Osteosarcoma. Nat. Rev. Dis. Primers 2022, 8, 77. [Google Scholar] [CrossRef] [Scilit]
- Moukengue, B.; Lallier, M.; Marchandet, L.; Baud’huin, M.; Verrecchia, F.; Ory, B.; Lamoureux, F. Origin and Therapies of Osteosarcoma. Cancers 2022, 14, 3503. [Google Scholar] [CrossRef] [Scilit]
- Cole, S.; Gianferante, D.M.; Zhu, B.; Mirabello, L. Osteosarcoma: A Surveillance, Epidemiology, and End Results program-based analysis from 1975 to 2017. Cancer 2022, 128, 2107–2118. [Google Scholar] [CrossRef] [Scilit]
- Huang, G.; Zhou, Y.; Wang, H.; Ma, X. Minimal residual disease and relapse surveillance in osteosarcoma: An action-linked framework integrating liquid biopsy and imaging biomarkers. J. Bone Oncol. 2026, 61, 100803. [Google Scholar] [CrossRef] [Scilit]
- Li, J.; Wang, A.; Guo, H.; Zheng, W.; Chen, R.; Miao, C.; Zheng, D.; Peng, J.; Wang, J.; Chen, Z. Exosomes: Innovative biomarkers leading the charge in non-invasive cancer diagnostics. Theranostics 2025, 15, 5277–5311. [Google Scholar] [CrossRef] [Scilit]
- Yu, D.; Li, Y.; Wang, M.; Gu, J.; Xu, W.; Cai, H.; Fang, X.; Zhang, X. Exosomes as a new frontier of cancer liquid biopsy. Mol. Cancer 2022, 21, 56. [Google Scholar] [CrossRef] [Scilit]
- Shin, H.; Choi, B.H.; Shim, O.; Kim, J.; Park, Y.; Cho, S.K.; Kim, H.K.; Choi, Y. Single test-based diagnosis of multiple cancer types using Exosome-SERS-AI for early stage cancers. Nat. Commun. 2023, 14, 1644. [Google Scholar] [CrossRef] [Scilit]
- Yu, W.; Hurley, J.; Roberts, D.; Chakrabortty, S.K.; Enderle, D.; Noerholm, M.; Breakefield, X.O.; Skog, J.K. Exosome-based liquid biopsies in cancer: Opportunities and challenges. Ann. Oncol. Off. J. Eur. Soc. Med. Oncol. 2021, 32, 466–477. [Google Scholar] [CrossRef] [Scilit]
- Preethi, K.A.; Selvakumar, S.C.; Ross, K.; Jayaraman, S.; Tusubira, D.; Sekar, D. Liquid biopsy: Exosomal microRNAs as novel diagnostic and prognostic biomarkers in cancer. Mol. Cancer 2022, 21, 54. [Google Scholar] [CrossRef] [Scilit]
- Wan, J.; Long, F.; Zhang, C.; Liu, Y. miR-181b-p53 negative feedback axis regulates osteosarcoma cell proliferation and invasion. Int. J. Mol. Med. 2020, 45, 1803–1813. [Google Scholar] [CrossRef] [Scilit]
- Jones, K.B.; Salah, Z.; Del Mare, S.; Galasso, M.; Gaudio, E.; Nuovo, G.J.; Lovat, F.; LeBlanc, K.; Palatini, J.; Randall, R.L.; et al. miRNA signatures associate with pathogenesis and progression of osteosarcoma. Cancer Res. 2012, 72, 1865–1877. [Google Scholar] [CrossRef] [Scilit]
- Han, Y.; Sun, X.; Cai, S. Advances in Research on Isothermal Signal Amplification Mediated MicroRNA Detection of Clinical Samples: Application to Disease Diagnosis. Biosensors 2025, 15, 395. [Google Scholar] [CrossRef] [Scilit]
- Wang, H. A Review of Nanotechnology in microRNA Detection and Drug Delivery. Cells 2024, 13, 1277. [Google Scholar] [CrossRef] [Scilit]
- Cheng, W.; Pan, H.; Chen, J.; He, M.; Wang, Z.; Xiang, Y. Enzyme-Free and Triple Sensitivity Amplification for Electrochemical Detection of Exosomal microRNA. Anal. Chem. 2025, 97, 5244–5250. [Google Scholar] [CrossRef] [Scilit]
- Qin, X.Y.; Dong, Z.Y.; Gao, F.; Huang, Y.Y.; Chen, Y.Y.; Chen, J.; Xie, B.P.; Duan, W.J.; Zhai, L.Y. A Simple and Rapid Bst Polymerase-Catalyzed Isothermal Exponential Amplification Reaction for Multiplex Detection of MicroRNAs. Anal. Chem. 2026, 98, 2840–2850. [Google Scholar] [CrossRef] [Scilit]
- Zhang, Y.; Sun, Y.; Gong, X.; Zhao, S.; Cao, W.; Wang, H.; Wang, H.; Zhang, X.; Du, C.; Chen, Z.; et al. Isothermal and protein-free cascade catalytic hairpin assembly induced-DNAzyme sensing strategy for sensitive miRNA analysis. Sci. Rep. 2025, 15, 44359. [Google Scholar] [CrossRef] [Scilit]
- Guo, J.; Zhao, S.F.; Chen, S.; Song, W.Q.; Li, L.Q. An enzyme-free electrochemical biosensor for sensitive and specific detection of microRNA-21 based on target recycling amplification and non-linear hybridization chain reaction. Sci. Rep. 2025, 15, 22272. [Google Scholar] [CrossRef] [Scilit]
- Yin, M.; Jiao, J.; Lu, L.; Hu, B.; Xue, L.; Dai, F.; Wang, X.; Wang, Z.; Wang, T.; Chen, Q. A simultaneous strategy with multiple-signal amplification and self-calibration for ultrasensitive assay of miRNA-21 based on 3D MNPs-IL-rGO-AuNPs. Biosens. Bioelectron. 2024, 249, 116009. [Google Scholar] [CrossRef] [Scilit]
- Zhang, X.; Li, W.; Zhou, Y.; Chai, Y.; Yuan, R. An ultrasensitive electrochemiluminescence biosensor for MicroRNA detection based on luminol-functionalized Au NPs@ZnO nanomaterials as signal probe and dissolved O2 as coreactant. Biosens. Bioelectron. 2019, 135, 8–13. [Google Scholar] [CrossRef] [Scilit]
- Zhang, W.; Song, Y.; Deng, D.; Liu, M.; Chen, H.; Zhang, W.; Lei, H.; Li, Z.; Luo, L. Exponential rolling circle amplification-hybridization chain reaction (EXRCA-HCR) for AgNPs@gel-enhanced fluorescence ultrasensitive detection of miRNA-21. Anal. Chim. Acta 2025, 1358, 344095. [Google Scholar] [CrossRef] [Scilit]
- Chen, H.; Zhuang, Z.; Chen, Y.; Qiu, C.; Qin, Y.; Tan, C.; Tan, Y.; Jiang, Y. A universal platform for one-pot detection of circulating non-coding RNA combining CRISPR-Cas12a and branched rolling circle amplification. Anal. Chim. Acta 2023, 1246, 340896. [Google Scholar] [CrossRef] [Scilit]
- Gong, S.; Zhang, S.; Wang, X.; Li, J.; Pan, W.; Li, N.; Tang, B. Strand Displacement Amplification Assisted CRISPR-Cas12a Strategy for Colorimetric Analysis of Viral Nucleic Acid. Anal. Chem. 2021, 93, 15216–15223. [Google Scholar] [CrossRef] [Scilit]
- He, X.; Deng, L.; Zhou, S.; Dong, J.; Zhu, S.; Li, J.; Li, X.; Huo, D.; Hou, C. CRISPR/Cas12a-coupled multiplexed strand displacement amplification for miRNA155 one-tube detection: Via a dual-cavity PCR tube. Mikrochim. Acta 2024, 191, 470. [Google Scholar] [CrossRef] [Scilit]
- Okumura, K.; Miyamoto, Y.; Mitarai, S.; Ato, M. One-step loop-mediated isothermal amplification system for Mycobacterium marinum detection. Microbiol. Spectr. 2025, 13, e0290624. [Google Scholar] [CrossRef] [Scilit]
- Zhang, M.; Wang, H.; Wang, H.; Wang, F.; Li, Z. CRISPR/Cas12a-Assisted Ligation-Initiated Loop-Mediated Isothermal Amplification (CAL-LAMP) for Highly Specific Detection of microRNAs. Anal. Chem. 2021, 93, 7942–7948. [Google Scholar] [CrossRef] [Scilit]
- Guo, H.; Chen, J.; Feng, Y.; Dai, Z. A Simple and Robust Exponential Amplification Reaction (EXPAR)-Based Hairpin Template (exp-Hairpin) for Highly Specific, Sensitive, and Universal MicroRNA Detection. Anal. Chem. 2024, 96, 2643–2650. [Google Scholar] [CrossRef] [Scilit]
- Qu, X.M.; Ren, X.D.; Su, N.; Sun, X.G.; Deng, S.L.; Lu, W.P.; Huang, Q. Isothermal exponential amplification reactions triggered by circular templates (cEXPAR) targeting miRNA. Mol. Biol. Rep. 2023, 50, 3653–3659. [Google Scholar] [CrossRef] [Scilit]
- Lin, Q.; Han, G.; Fang, X.; Chen, H.; Weng, W.; Kong, J. Programmable Analysis of MicroRNAs by Thermus thermophilus Argonaute-Assisted Exponential Isothermal Amplification for Multiplex Detection (TEAM). Anal. Chem. 2022, 94, 11290–11297. [Google Scholar] [CrossRef] [Scilit]
- Rahn, K.L.; Peramune, U.; Zhang, T.; Anand, R.K. Label-Free Electrochemical Methods for Disease Detection. Annu. Rev. Anal. Chem. 2023, 16, 49–69. [Google Scholar] [CrossRef] [Scilit]
- Wang, Z.; Chen, Z.; Ma, Z.; Han, H. Label-Free Mode Based on Ferrocene/PEDOT:PSS-PPy for Molecularly Imprinted Electrochemically Ultrasensitive Detection of Amino Acids. Anal. Chem. 2024, 96, 14298–14305. [Google Scholar] [CrossRef] [Scilit]
- Bo, B.; Zhang, T.; Jiang, Y.; Cui, H.; Miao, P. Triple Signal Amplification Strategy for Ultrasensitive Determination of miRNA Based on Duplex Specific Nuclease and Bridge DNA-Gold Nanoparticles. Anal. Chem. 2018, 90, 2395–2400. [Google Scholar] [CrossRef] [Scilit]
- Sage, A.T.; Besant, J.D.; Lam, B.; Sargent, E.H.; Kelley, S.O. Ultrasensitive electrochemical biomolecular detection using nanostructured microelectrodes. Acc. Chem. Res. 2014, 47, 2417–2425. [Google Scholar] [CrossRef] [Scilit]
- Barman, S.C.; Ali, M.; Hasan, E.A.; Wehbe, N.; Alshareef, H.N.; Alsulaiman, D. Smartphone-Interfaced Electrochemical Biosensor for microRNA Detection Based on Laser-Induced Graphene with π–π Stacked Peptide Nucleic Acid Probes. ACS Mater. Lett. 2024, 6, 837–846. [Google Scholar] [CrossRef] [Scilit]
- Moccia, M.; Caratelli, V.; Cinti, S.; Pede, B.; Avitabile, C.; Saviano, M.; Imbriani, A.L.; Moscone, D.; Arduini, F. Paper-based electrochemical peptide nucleic acid (PNA) biosensor for detection of miRNA-492: A pancreatic ductal adenocarcinoma biomarker. Biosens. Bioelectron. 2020, 165, 112371. [Google Scholar] [CrossRef] [Scilit]
- Abdel-Hamied, M.; Guo, M.; Wei, Y.; Bansmann, J.; El Nashar, R.M.; Oswald, F.; Mizaikoff, B.; Kranz, C. A novel hybrid biosensor for miRNA detection based on peptide nucleic acids and molecularly imprinted polymers. Bioelectrochemistry 2025, 165, 108964. [Google Scholar] [CrossRef] [Scilit]
- Wang, B.; Zhou, S.; Zhang, X.; Wang, R.; Huang, S.; Li, A.; He, H.; Zhang, Y.; Wei, Y.; Yang, Z.; et al. PRICE: Direct and robust detection of microRNAs at single-nucleotide resolution. Nat. Commun. 2026, 17, 2647. [Google Scholar] [CrossRef] [Scilit]
- Li, Z.; Lv, Y.; Duan, X.; Liu, B.; Zhao, Y. Highly Uniform DNA Monolayers Generated by Freezing-Directed Assembly on Gold Surfaces Enable Robust Electrochemical Sensing in Whole Blood. Angew. Chem. (Int. Ed. Engl.) 2023, 62, e202312975. [Google Scholar] [CrossRef] [Scilit]
- Ma, C.; Zhou, Q.; Shi, J.; Gao, H.; Huang, D.; Xue, H.; Wang, H.; Zhang, Z.; Yang, S.; Zhang, J.; et al. CRISPR-empowered electrochemical biosensor for target amplification-free and sensitive detection of miRNA. Talanta 2024, 266, 125125. [Google Scholar] [CrossRef] [Scilit]
- Niu, C.; Liu, J.; Xing, X.; Zhang, C. CRISPR-Cas12a-assisted elimination of the non-specific signal from non-specific amplification in the Exponential Amplification Reaction. Anal. Chim. Acta 2023, 1251, 340998. [Google Scholar] [CrossRef] [Scilit]
- Dong, H.; Huang, R.; Yang, D.; Zhao, J.; Lin, B.; Pan, Y.; Lin, X.; Yang, Y.; Guo, Z.; Li, N.; et al. Just-in-Time Generation of Nanolabels via In Situ Biomineralization of ZIF-8 Enabling Ultrasensitive MicroRNA Detection on Unmodified Electrodes. Anal. Chem. 2024, 96, 16793–16801. [Google Scholar] [CrossRef] [Scilit]
- Sun, M.; Zhang, Y.; Xie, J.; Zhang, Y.; Huang, T.; Li, M.; Chen, J.-X.; Dai, Z.; Chen, J. An APE1-mediated isothermal target cycling amplification for label-free and rapid detection of miRNA. Sens. Actuators B Chem. 2024, 401, 135029. [Google Scholar] [CrossRef] [Scilit]
- Liu, Y.; Li, Y.; Gong, H.; Liu, Y.; Wang, Y.; Ma, C.; Wei, Y.; Shi, C. A rapid, specific, and simple-to-use biosensor for amplification-free determination of microRNA based on electrical potential-assisted and ternary hybridization. Mikrochim. Acta 2025, 192, 315. [Google Scholar] [CrossRef] [Scilit]
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |
© 2026 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license.





