Introduction

MicroRNA (miRNA) is recognized as a critical cancer marker [1, 2]. The miRNA-21 is strongly associated with colorectal cancer, breast cancer, lung cancer, and pancreatic cancer, as well as cervical cancer [3,4,5,6,7]. However, the miRNA-21 level in the human body is low. Hence, achieving a sensitive determination of miRNA-21 is important for clinical analysis and diagnosis of cancer. The usual ways to detect miRNA-21 are based on the Northern Blot and qRT-PCR [8, 9]. Despite the high sensitivity of these methods, they rely on complex equipment and require a rigorous experimental environment. This may restrict the further application of them for portable and on-site miRNA determination.

To achieve the portable determination, electrochemical techniques were wildly developed in recent years owing to the benefits like simple operation, effective response, and strong adaptability [10, 11]. However, there are two difficult requirements for the sensitive determination of miRNA-21, which includes a stabilized beacon and an enzyme-free signal amplification strategy. The general electrochemical beacons such as methylene blue, ferrocene, and thionine are easily influenced by modification status, interface condition, and solution pH, which may cause false negative or positive results. Accordingly, the nanozymes made by metal elementary substances or compounds and possess catalytic property are proposed [12]. For example, Xia et al. developed multiple Ag NPs–based electrochemical biosensors and strategies for the detection of miRNAs and other biomarkers [13, 14]. Additionally, metallic oxides like Fe3O4 NPs and CuO NPs, as well as some carbon-based nanomaterials, also possess favorable catalytic activity and strong stability, and can generate a measurable electrochemical signal with H2O2 [15,16,17,18]. Nonetheless, the catalysis levels of single metal nanozymes are constrained. For improved performance, bimetallic nanozymes with heterostructures or homostructures have been developed. The bimetallic nanozymes possess larger specific surface area and continuous ion exchange capability, resulting in a shorter interaction time and stronger electron transfer level [19,20,21], which provides a solid supporting for high-efficient catalysis.

As for signal amplification strategies, traditional methods based on the polymerase chain amplification technique were limited by enzymatic activity and the ion effect [22]. With the development of DNA self-assembling, multiple enzyme-free amplification ways were proposed like DNA train, DNA branch, and DNA polymer [23, 24]. However, the DNA strand cannot infinite assembling owing to the steric hindrance effect [25], and the unspecific DNA adsorption commonly causes an unreliable result. In comparison of this direct assembling, a new amplification way which depends on toehold-mediated DNA chain replacement was proposed. Toehold is a short oligonucleotides; it can accurately recognize targets, and orderly release the complementary sequence to perform a cyclical DNA assembling for signal amplification. Currently, the toehold technique provides a solid support for trace biomarker sensitive determination [26,27,28,29].

Herein, in consideration of catalytic ability, biocompatibility, and metallic property, Au and Pt are chosen to prepare AuPt nanoparticles (AuPt NPs) as catalytic beacon. A sensitive determination of miRNA-21 in serum has been established in synergy with a toehold-mediated DNA replacement strategy. Once the miRNA-21 is recognized by the toehold sequence, a continuous cyclic chain replacement can be performed, resulting in the accumulation of the AuPt NPs onto electrode surface. In the presence of H2O2, a measurable electrochemical signal is obtained via the catalytic interaction. Different from other electrochemical methods based on nanozymes, the proposed method utilizes a bimetallic nanozyme that interacts with the cyclic chain substitution effect to achieve selective and sensitive measurements.

Experimental section

Reagents and apparatus

L-ascorbic acid (AA), chloroauric acid (HAuCl4), chloroplatinic acid (H2PtCl6), and oligonucleotides (oligos) were obtained from Sangon Biotech Co. Ltd. (Shanghai, China); the oligos sequence listed in Table S1. Trimethylol aminomethane (Tirs), Tris (2-carboxyethyl) phosphine hydrochloride (TCEP), L-proline, 6-mercaptohexanol (MCH), 3,3′,5,5′-tetramethylbenzidine (TMB), 2,2′-azino-bis-(3-ethylbenzthiazoline- 6-sulfonic acid) (ABTS), and o-phenylenediamine (OPD) were purchased from Aladdin (Shanghai, China). Na2HPO4, NaH2PO4, K3Fe(CN)6, K4Fe(CN)6, and H2O2 were obtained from Guangzhou Chemical Reagent Factory (Guangzhou, China). The other reagents utilized in the experiments were analytical grades. The ultrapure water was obtained through Milli-Q® Integral (Milli-Q, France).

The oligos sequence is designed with an assist of ensemble website according to the complementary effect and toehold replacement requirements. The electrochemical workstation with a three-electrode system including a gold working electrode (3 mm in diameter), an Ag/AgCl reference electrode, and a platinum wire counter electrode belongs to Metrohm Autolab PGSTAT302N (Eco Chemie, Netherlands). This device is utilized to make electrochemical assays. The UV–Vis spectrum assays were performed by UV-2700 ultraviolet spectrophotometry (Shimadzu, Japan). The surface diffraction assays of the prepared nanozymes were performed by Ultima IV X-ray Powder Diffractometer (Rigaku, Japan) and nanoparticle size potential analyzer which belongs to Marvin Zetasizer Nano ZS90 (Malvern, Britain). In the assays, the 1.0 mg of AuPt NPs was dispersed into 1.0 mL H2O by ultraphonic for the XPS, the 1.0 μM H0, 1.0 mg/mL AuPt NPs, and 1.0 mg/mL of H0@AuPt NPs were prepared for Z-potential test respectively, and the 0.02 mg/mL of AuPt NPs was collected for the EDX test. The morphology and element distribution characterizations were performed by FEI Talos F200X field emission transmission electron microscope (FEI, USA) and Helios G4 CXTalos scanning electron microscope (Zeiss, Germany).

Preparation of AuPt NPs

The preparation of AuPt NPs is based on the procedure of the reported information [30]. The detailed information is shown in Supporting Information.

Enzymatic activity analysis of AuPt NPs

The enzymatic activity is assessed via the value of Km. The detailed information is shown in Supporting Information.

In addition, the level of colorimetric response was investigated by spectroscopic tests. Taking 1.5 mL of 0.2 M HAc-NaAc buffer (pH 3.5), 200 μL of 0.4 mM H2O2, 200 μL of 20 mM TMB, and 100 μL of 1 mg/mL AuPt NPs to mix and react at 37℃ for 5 min. UV–Vis absorption spectra in the wavelength range from 350 to 800 nm were recorded. Subsequently, the TMB was replaced by ABTS and OPD respectively to repeat the test.

Preparation of DNA@AuPt NPs

The 1.0 mg of AuPt NPs was dispersed into 1.0 mL of 0.1 M PBS (pH 7.4) by ultraphonic, and the 40 μL of 10 μM SH-DNA that activated by TCEP was added into 400 μL of the PBS containing AuPt NPs. The mixture was stirred at 4℃ for 30 min, and then centrifuged at 10,000 rpm for 5 min. The 20 mM Tris–HCl buffer (pH 7.4) contained 100 mM NaCl and 5 mM MgCl2 was utilized to wash the product two times. The final product was stored in 400 μL of 20 mM Tris–HCl buffer (pH 7.4) for further use.

Preparation of DNA sensor

Gold electrode (GE) was firstly immersed in piranha solution (30% H2O2:98% H2SO4 = 3:7/v:v) for 30 min, and rinsed by ultrapure water. It was then polished to a mirror-like surface by Al2O3 powder with 1.0, 0.3, and 0.05 µm diameter, respectively. Subsequently, an activation test was performed through a 10 cycle of scanning in 50 mM H2SO4 solution (− 0.2 ~ 1.6 V).

Taking 1 mL of 20 mM Tris–HCl buffer (pH 7.4) contained 100 mM NaCl and 5 mM MgCl2, 1 μM SH-DNA (H1), 1.5 μM complementary strand (H2), and 1.5 μM switch strand (H3) to make DNA hybridization at 37℃ for 3 h. The hybridized double-strand DNA (dsDNA) was then activated by TCEP. Subsequently, the 30 μL of the activated dsDNA was dropped onto the polished GE surface, and stored at 4℃ overnight. After washing by Tris–HCl buffer (pH 7.4), a 40 μL of 1.0 μM MCH was dropped onto the GE surface, and incubated at 4℃ for 30 min. After washing by the Tris–HCl buffer, the DNA sensor (GE/MCH/dsDNA) was formed.

Determination of miRNA-21

The 30 μL of 20 mM Tris–HCl buffer (pH 7.4) contained 100 mM NaCl and 5 mM MgCl2, as well as different concentrations of DNA@AuPt NPs and miRNA-21 was taken to the surface of the prepared DNA sensor to incubate at 37℃ for 3.5 h. After washing by 0.1 M PBS buffer (pH 7.4), the sensor is then immersed in 10 mL of the PBS buffer to perform “i-t” assays with the addition of H2O2. The initial potential is recorded at − 0.4 V, and after addition of 10 μL of 5 M H2O2, a gentle stirring step with 5 s is performed. Before the electrochemical measurements, the stirring is stopped.

Practical assays

The serum sample obtained from the Second People’s Hospital of Guangdong was filtered by a membrane with 0.2 μm diameter. Subsequently, they were extracted by a miRNA kit purchased from TianGen biochemical technology limited company. After the extraction, the miRNA contained in the serum sample was collected and transferred into the test tank with PBS buffer (pH 7.4). The electrochemical assays were then continued as before. In the spiking assays, different concentrations of miRNA-21 were respectively spiked in the serum sample. Similarly, they were filtered by a membrane with 0.2 μm diameter to perform the tests as before.

Results and discussion

Principle

The principle of the DNA sensor for the determination of miRNA is revealed in Scheme 1. Firstly, a SH-DNA (H1) is hybridized with a complementary strand (H2) and a switch strand (H3). Meanwhile, another SH-DNA (H0) is modified onto the AuPt NPs to form a beacon (H0@AuPt NPs). The H1 has performed a probe role which contained a toehold sequence, the H2 provides a supporting effect, and the H3 can avoid the combining of H0@AuPt NPs to the H1. Nonetheless, in the presence of miRNA-21, the H3 is released via toehold-mediated DNA chain replacement effect. The generated sequence change further led to a secondary replacement between the H0@AuPt NPs and the miRNA-21, and the released miRNA-21 is constrained to replace another H3. This results in a continuous cyclic replacement, which progressively accumulates plenty of H0@AuPt NPs on the electrode surface. While transferring the electrode into PBS buffer with H2O2, a redox was then catalyzed, which caused a measurable electrochemical response.

Scheme 1
scheme 1

The principle of the electrochemical DNA sensor for the detection of miRNA-21. (A) Preparation of the beacon (H0@AuPt). (B) Electrochemical detection procedure for miRNA-21

Full size image

Characterization of AuPt NPs

To demonstrate the successful synthesis of AuPt NPs, TEM, SEM, and EDS are employed to investigate the morphology and constitution of the product. As the TEM graph displayed (Fig. 1A and B), the product presents a spherical structure with a diameter of ~ 50 nm. The following SEM and EDS graphs (Fig. 1C ~ G) indicated that the product contained Au and Pt elements with a uniformed distribution, and the atomic ratio of Au and Pt was 59:41. It is worth noting that Au and Pt coexist during the preparation procedure. The PtCl62− and AuCl4− precursors reduce to metallic Pt and Au atoms with the addition of AA, and fuse together to form AuPt nuclei. Afterwards, the AuPt nuclei gradually extended to larger nanoparticles by L-proline [30].

Fig. 1
figure 1

Characterization of the prepared AuPt NPs. (A, B) TEM images of different sizes of the AuPt NPs. (C–G) EDS images of the AuPt NPs. The C indicates the scanning area, and the D exhibits the mixture diagram of Au and Pt elements

Full size image

To further study the elemental composition and valence state of the AuPt NPs, a XPS assay is performed. In the full spectrum (Figure S1A), it contained Au and Pt, as well as C, N, and O elements which are from the sample substrate and air. In the high-resolution Au 4f XPS spectrum (Figure S1B), it shows two specific peaks at 86.68 eV and 83.04 eV; they belong to Au 4f 5/2 and Au 4f 7/2 in Au0 state, suggesting the Au precursor is reduced [31]. In the high-resolution Pt 4f XPS spectrum (Figure S1C), two pairs of peaks of Pt 4f 5/2 and Pt 4f 7/2 are observed. The peaks at 73.48 eV and 70.18 eV correspond to Pt0 state, indicating that Pt4+ is reduced. Another peak pair at 75.18 eV and 71.48 eV are derived from Pt2+, which are obviously lower than those of the Pt0, indicating the Pt0 is the dominant [32]. Besides, in comparison of the standard Pt peak, the peaks of the Pt 4f appeared displacement. This is because the electron interaction between Au and Pt can change the d-band structure of Pt, which is beneficial to performance and stability of the material [33, 34]. Therefore, the AuPt NPs are proved to be synthesized. Additionally, to verify the connection of AuPt NPs and DNA (H0), a Zeta electrical potential assay is made. Figure S1D shows that the Zeta potentials of the H0 and AuPt NPs are − 4.9 and − 27.8 mV. After modification, the Zeta potential is changed to − 32.8 mV, indicating the negatively charged DNA is covered on AuPt NPs to form the H0@AuPt NPs.

DNA sensor construction and performance test

Electrochemical impedance is utilized to illustrate the construction procedure of the DNA sensor. The sensors with different modification levels are respectively incubated into [Fe(CN)6]3−/4− solution to observe the signal. As Fig. 2A shows, little impedance signal is observed for bare electrode (line a), and the signal is orderly increased after the modifications of H1 ~ H3 (line b) and MCH (line c). It is worth noticing that the assembling of the electroactive material H0@AuPt NPs to H1 led to the increase of the impedance signal (line d). This is because the hybridized DNA strands between the AuPt NPs and electrode are not complete complementary; the mismatch bases led to a limited electron transfer. This phenomenon is proved by some reported literature [35, 36], whereas this cannot influence the signal response produced by the redox in the buffer. The sensor assembling procedure is additionally confirmed by the cyclic voltammetry assays in [Fe(CN)6]3−/4− solution (Figure S2). The CV signals (Fe(III)/Fe(II)) exhibited a reversed trend to that of the impedance, suggesting the modification is expected. Meanwhile, the SEM graphs (Figure S3) for the electrodes before and after the modifications further evidenced the construction of the DNA sensor.

Fig. 2
figure 2

(A) Nyquist plots of EIS to the [Fe(CN)6]3-/4- solution containing 0.1 M KCl (a) bare Au electrode, (b) Au electrode/H1/H2/H3, (c) Au electrode/H1/H2/H3/MCH, and (d) Au electrode/H1/H2/H3/MCH/H0@AuPt. (B) Lineweaver–Burk plot when the TMB concentration is fixed as 2.0 mM. (C) The ratio of current responses of the DNA sensor with toehold sequence to different concentrations of miRNA-21 (pM) (line a for 0, line b for 1.0, line c for 100). (D) The ratio of current responses of the DNA sensor with toehold sequence (line a) to 0 pM miRNA-21, as well as that without (line b) and with (line c) toehold sequence to 100 pM miRNA-21

Full size image

The following assays focus on the signal response of the sensor. First, an i-t test is used to assess the sensor response in the absence of the target miRNA. As Figure S4A shows, a bare electrode (line a) displays little current response. However, after the assembling of the H0@AuPt NPs to the DNA sensor via toehold DNA replacement effect, it can result an obvious current signal (line b) with the addition of H2O2. While if the H2O2 is absent in the test solution, the current signal is little (line c) like that produced by the bare electrode. This indicates that the AuPt NPs has a satisfactory catalytic property for the electrochemical determination. Moreover, the colorimetric assays further indicate that AuPt NPs are capable of catalyzing H2O2 to generate ·OH, and results in the oxidation of TMB, ABTS, and OPD to appear color conversions (Figure S4B). In this case, the enzymatic activity is then investigated via the kinetic process and the corresponding Km. The assays are performed at 25℃, and the pH of the test buffer was controlled at 3.5. According to the obtained kinetic curve and Lineweaver–Burk plot (Fig. 2B), the Km value is calculated as 0.072 mM for H2O2, which is better than that of the horseradish peroxidase which is 3.7 mM [37, 38].

Encouraged by the above results, an assay for miRNA is performed. Figure 2C illustrates the i-t response of the DNA sensor containing toehold sequence to miRNA-21 at concentrations of 0, 1.0, and 100 pM. To perform the comparison more definitely, the responses are displayed by current ratio, which the level of the max current response is normalized as “1,” and the other current responses are shown according to the intensity ratio. The current responses to different concentrations of miRNA-21 exhibit measurable distinctions, suggesting a favorable determination feasibility of the DNA sensor for miRNA-21. In Fig. 2D, it indicates the responses of the DNA sensor with toehold sequence to 0 (line a) and 100 (line c) pM miRNA-21, as well as that without toehold sequence to 100 (line b) pM miRNA-21. The responses are displayed by current ratio like in Fig. 2C. The level of line b is within the fluctuation range of line a, suggesting the DNA sensor without toehold sequence is unable to recognize miRNA-21, and incapable to lead the assembling of H0@AuPt NPs. However, the response distinctions of the DNA sensor with toehold sequence to 0 and 100 pM miRNA-21 are obvious. This proves that the toehold sequence provides solid support for target recognition and signal amplification.

Determination of miRNA

To effectively measure the level of target miRNA, a condition optimization assay was initially performed. Firstly, the amount of the H0 modifying on AuPt NPs is investigated. As Fig. 3A shows, once the modified concentration arrived at 1 µM, the signal response is the largest. When the H0 amount continuously increased, the signal is declined owing to the steric-hinerance effect preventing further chain replacements. Similarly, the amount of the H1 modifying on the electrode is then assessed. As Fig. 3B shows, the 1 µM is the optimal usage. Subsequently, the incubation time at 37℃ is investigated. As Fig. 3C shows, when the time is controlled at 3.5 h, the signal is the most obvious, while the longer incubation can cause the deposition of H0@AuPt NPs, producing more background signal, and decreasing the specific signal distinction.

Fig. 3
figure 3

Optimization of experimental conditions. (A) Modification concentration of H0 on the AuPt NPs. (B) Immobilization concentration of the H1 probes on the Au electrode. (C) Incubation time of H0@AuPt probes

Full size image

Under the optimal conditions, the electrochemical miRNA determinations are performed. As Figure S5 depicts, the relative current conversions ((I-I0)/I0) are gradually increase with the miRNA amount raising from 0.1 pM to 5 nM, and a linearity is plotted by the ((I-I0)/I0) versus the logarithmic concentrations (lgC) of 0.1 ~ 1000 pM (Fig. 4A). The fitted linear equation is (I-I0)/I0 = 0.2167lgC + 0.5147, the LOD is calculated as 84.1 fM (3σ/S), and the analytical range is confirmed from 0.1 pM to 1.0 nM. In comparison of other miRNA determination methods, this strategy has preferable benefits on the analytical range and LOD (Table S2). Subsequently, the selectivity of the strategy is evaluated. Four isogenous miRNAs of miRNA-141, miRNA-101, miRNA-122, miRNA-let7a are chosen as the interference; the concentrations of them are 100 pM, which is 100-fold to that of miRNA-21. As Fig. 4B reveals, solely miRNA-21 causes an obvious signal response, and the mixture containing the miRNA-21 and other four isogenous miRNA present a similar signal, which exhibits a satisfactory selectivity of this strategy. In addition, five prepared DNA sensors are respectively used to detect 1.0 pM miRNA-21, and the assays are repeated 3 times (n = 3). According to the results, the rates of the signal conversion of the electrodes are similar, and the RSD value from each electrode is under 3.0% (Fig. 4C). Moreover, the DNA sensors were then stored at 4℃ for 1, 2, 3, 4, and 5 days, and employed to perform the miRNA-21 assays. The results demonstrated the signals remaining at a similar level (Fig. 4D). Consequently, the favorable reproducibility and stability of the proposed strategy were verified.

Fig. 4
figure 4

(A) The working curve of i-t responses to target miRNA-21. (B) Investigation of the specificity of the sensor, the relative current change ratios responding to blank, miRNA-141, miRNA-101, miRNA-122, miRNA-let7a, miRNA-21, and their mixture. (C) Investigation of the repeatability of the sensor. (D) Investigation of the stability of the biosensor

Full size image

To assess the practicability of the proposed eletcrochemical strategy, it is utilized to make the miRNA-21 determination in serum. The serum sample is filtered by a membrane (0.2 μm diameter) to dislodge impurities, and extracted by a commercial miRNA kit. The pre-treated serum sample is mixed with 20 mM Tris–HCl buffer (pH 7.4) to make measurement. Furthermore, different concentrations of miRNA-21 are respectively spiked into the original serum sample; after filtration and extraction, the product is then perform the measurements. As Table 1 displays, the miRNA-21 is not detected in the serum sample, and the spiked recoveries of the 1, 10, and 100 pM miRNA-21 are 95.0%, 101.5%, and 102.4% with the RSD values ranged from 3.7 to 5.8%. This indicated that the proposed electrochemical determination strategy owns a desirable application prospect.

Table 1 Recovery test of spiked samples by the proposed determination strategy
Full size table

Conclusion

This work proposed a sensitive electrochemical determination strategy for the miRNA-21 in serum. A bimetallic AuPt nanozyme possessing favorable peroxidase property is prepared as beacon, and a continuous toehold-mediated DNA replacement procedure is employed to make sensitive determination of miRNA-21. After addition of few H2O2, a measurable signal to miRNA-21 amount is obtained, the sensitive arrived at femtomole. By contrast to common miRNA-21 determination methods, this strategy has multiple benefits: achieving enzyme-free signal amplification and dispensing with high-cost instrument, as well as facilitating field detection. These indicate a promising application prospect in clinical testing and disease prediction.