Cascade Toehold-Mediated Strand Displacement Reaction for Ultrasensitive Detection of Exosomal MicroRNA

At a Glance

Metadata	Details
Publication Date	2020-10-13
Journal	CCS Chemistry
Authors	Peng Miao, Yuguo Tang
Institutions	Suzhou Institute of Biomedical Engineering and Technology, New York University
Citations	33

Abstract

Open AccessCCS ChemistryCOMMUNICATION1 Jul 2021Cascade Toehold-Mediated Strand Displacement Reaction for Ultrasensitive Detection of Exosomal MicroRNA Peng Miao and Yuguo Tang Peng Miao Corresponding author: E-mail Address: [email protected] Suzhou Institute of Biomedical Engineering and Technology, Chinese Academy of Sciences, Suzhou, Jiangsu 215163 Department of Chemistry, New York University, New York, NY 10003 and Yuguo Tang Suzhou Institute of Biomedical Engineering and Technology, Chinese Academy of Sciences, Suzhou, Jiangsu 215163 https://doi.org/10.31635/ccschem.020.202000458 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesTrack Citations ShareFacebookTwitterLinked InEmail MicroRNA (miRNA) in exosomes is a powerful molecular signature for early diagnosis of cancers with the merits of high specificity and high stability. Herein, we report an ultrasensitive electrochemical assay to measure miRNA using a cascade toehold-mediated strand displacement reaction (SDR). In SDR, the trapped exosomal miRNA releases a large amount of single-stranded DNA in the solution. The product then triggers the downstream SDR at the electrode surface. Furthermore, reductant-assisted electrochemical amplification is introduced to adjust the recorded signal intensity. Under optimal conditions, ultrahigh sensitivity is achieved with excellent signal amplification efficiency. The proposed method also performs well in biological samples and successfully distinguishes patient samples from healthy controls in practical use. Therefore, this work provides a promising approach for exosomal miRNA analysis, which has great potential utility in cancer diagnostics. Download figure Download PowerPoint Introduction Exosomes are a typical extracellular vesicles species secreted by various parent cells, which contain abundant specific RNAs, proteins, cytokines, and so on.1-3 They exist in different body fluids like urine and blood.4-6 These samples can be easily obtained. Circulating exosomes have central roles in many important physiological or pathological processes by transporting cargoes originating from parent cells.7,8 Among them, microRNAs (miRNAs), a type of endogenous single-stranded and noncoding RNA, are implicated in tumor progression by participating in microenvironment modulation, immune evasion, and angiogenesis.9,10 Accumulated evidence has demonstrated miRNA as a reliable indicator for early diagnosis of diseases.11 For instance, miR-21 is highly enriched in different tumor exosomes (e.g., lung, ovarian, brain, and breast cancers).12 In addition, exosomal miRNA can be well protected by the outer lipid membrane of exosomes against degradation by RNase. Due to the easy acquisition of noninvasive samples and high stability of target, exosomal miRNA has huge potential as an optimal disease biomarker.13 In the past, quantitative reverse transcription polymerase chain reaction (qRT-PCR) has been regarded as the gold standard for the quantification of miRNAs with fM sensitivity.14 However, traditional qRT-PCR requires complicated enzymatic amplification, large sample volume, and tedious primer designs, which can be easily contaminated.15 Recently, novel approaches have been developed based on the techniques such as fluorescence assay,16 surface-enhanced Raman scattering (SERS),17 surface plasmon resonance (SPR),18 mass spectrometry,19 sequencing,20 and so on. Nevertheless, in most of these methods, expensive instruments and complicated operations hinder their extensive applications. Electrochemistry is an excellent alternative technique for analytical purposes.21-24 A few electrochemical methods have already been established to measure exosomal miRNA levels. For example, Liu et al.25 integrated neutrally charged peptide nucleic acid (PNA) and nucleic acid-modified gold nanoparticles for enhanced detection of tumor exosomal miRNA by chronocoulometry. Boriachek et al.26 prepared DNA-modified magnetic beads for the capture and amplification-free measurement of the target. Guo et al.27 achieved DNA elongation on the electrode interface by hybridization chain reaction and determined exosomal miRNA levels with reduced false-positive signals. Toehold-mediated strand displacement reaction (SDR) is an ingenious strategy to fabricate and adjust DNA structures.28 The reaction is nonenzymatic and entropy driven, which is based on the variation of free energy and the principle of toehold exchange. At first, the miRNA target usually hybridizes with a short DNA overhang domain (toehold) at a partially hybridized duplex. The length is short, but the initiation of the subsequent hybridization reaction is greatly accelerated. By careful design, the displacement process is followed by the formation of a DNA duplex, which is fast and predictable. Meanwhile, mismatched sequences can be easily differentiated, and high sensitivity and selectivity are promised. Using these considerations, we herein propose a novel cascade toehold-mediated SDR strategy and fabricate a simple but ultrasensitive electrochemical method to measure exosomal miRNA. Six DNA probes with multiple functional regions are prehybridized or immobilized and then used for reactions in homogeneous phase or at the solid interface, respectively. This method shows several unique features. First, in target-triggered upstream SDR, the target and DNA strand that initiates downstream SDR can be displaced simultaneously, therefore cascade SDR can be well integrated. Second, reductant-mediated amplification is applied, and the whole system is enzyme-free, thus promising a robust practical use. Third, the presence of miRNA leads to a conformation change of the DNA monolayer on the electrode, which can be facilely measured by the response of an electrochemical species. Compared with some other nonenzymatic amplifications, like the hybridization chain reaction, the distance between electrochemical species and electrode is not only closer but also fixed, which is more suitable for quantitative analysis.27 Experimental Methods Modification of working gold electrode Substrate electrode was first pretreated using piranha solution (98% H2SO4:30% H2O2 = 3∶1) for 5 min (Caution: Piranha solution dangerously attacks organic matter!). After rinsing, polishing was carried out using silicon carbide paper (P5000) and alumina powders. Then, the electrode was sonicated in ethanol and double-distilled water, successively. Next, it was immersed in nitric acid (50%) for 30 min and then electrochemically cleaned with H2SO4 (0.5 M). After drying with nitrogen, the cleaned electrode was treated with probe E (0.7 μM, 10 mM Tris-HCl, 10 mM tris(2-carboxyethyl)phosphinehydrochloride (TCEP), 0.1 M NaCl, 1 mM EDTA, pH 7.4) for 8 h, followed by incubation with 6-mercaptohexanol (1 mM) for 30 min. Fabrication of the miRNA sensor Solutions of probes A, B, and C were prepared at a concentration of 0.18 μM with 10 mM Tris-HCl buffer. After mixing to a ratio of 1∶1∶1, the solution was heated to 95 °C for 300 s and then cooled slowly to 25 °C. After that, samples containing different concentrations of miRNA and 0.06 μM probe D were blended with the above triple-stranded substrate. After reaction for 30 min, 0.06 μM probe F was added, and probe E-modified electrode was treated with the above solution for 30 min before electrochemical measurements. Electrochemical detection A traditional three-electrode system was used, consisting of gold working electrode, platinum auxiliary electrode, and saturated calomel reference electrode. Electrochemical impedance spectroscopy (EIS) was carried out in electrolyte A {5 mM [Fe(CN)6]3−/4−, 1 M KNO3}. Square wave voltammetry (SWV) was carried out in electrolyte B (20 mM Tris-HCl, 160 μM TCEP, 50 mM MgCl2, 100 mM NaCl, and pH 7.4). The EIS was conducted at 0.22 V biasing potential, 5 mV amplitude, and the 0.1 Hz to 100 kHz frequency range. In contrast, the SWV was obtained at a 50 mV/s scan rate, 4 mV step potential, 70 Hz frequency, and −0.05 to −0.55 V scan range. Exosomes isolation and RNA extraction Centrifugation was used for exosomes isolation and commercial kit was applied for RNA extraction.29,30 The cells were cultured in DMEM containing 10% fetal bovine serum. After the cells reached exponential growth phase, they were first washed with phosphate-buffered saline (PBS) and then cultured with serum-free DMEM for another 2 days. The culture supernatants were centrifuged at 2000g (20 min), 10,000g (30 min), and 110,000g (120 min), successively. The exosomes were thus isolated. RNA extraction was achieved by employing the TRIzol total RNA isolation kit. A brief protocol was as follows: Lysate was first spiked with the exosomes. Then, DNA and protein complexes were removed after centrifugation. After adding deproteinized liquid, the samples were centrifuged for the second time. Next, rinse liquid was added. After centrifugation, RNA was harvested and resuspended in RNase-free water. Detection of exosomal miRNA from patient samples Blood samples from healthy individuals and breast cancer patients were collected from Suzhou Science & Technology Town Hospital (Suzhou, China), which was approved by the Ethics Committee of the hospital. All people gave informed consent before collection. The samples were first pretreated with proteinase K (60 μg/mL) and RNase A (4 U/mL) to eliminate contaminations of proteins and RNA complexes. Then, the samples were subjected to exosomal miRNA extraction and electrochemical measurements. Results and Discussion The sensing mechanism can be easily understood by the illustration in Scheme 1. miR-21 is employed as the model target. It is first isolated from tumorigenic exosomes. For upstream SDR, probes A (c, b*, a*, d*, g*, f*, and e*), B (e, f, and g), and C (d and a) in the ratio of 1∶1∶1 are mixed to form the triple-stranded substrate, which contains a single-stranded region partially complementary with the target. The toehold domain of c* and b* pairs with part of miR-21 (b and c), and the rest of the target (a) replaces the corresponding region of probe C, which is liberated, and a new toehold (d*) is left in the middle of DNA substrate. Next, probe D (f, g, d, a, and b) pairs with the new toehold and then replaces the probe B domain (f and g) and miR-21 domain (a and b). These two strands are thus displaced with a gain in configurational entropy. The recycled target triggers another round of upstream SDR, thus a minimal amount of target is required to generate sufficient probe B for the initiation of downstream SDR. The cascade SDR leads to significant electrochemical response variation, which is suitable for the determination of exosomal miRNA. Briefly, hairpin-structured probe E (g*, f*, e*, g, h, e, f, and i) with a thiol modification at the 5′ terminal is modified on the surface of the electrode. Methylene blue (MB) labeled at the 3′ terminal is used to provide electrochemical signal. Also, we apply TCEP to enhance the initial electrochemical signal.31 Oxidized MB is reduced by TCEP and the cyclically activated MB contributes to significantly amplified electrochemical response. The product of upstream SDR (probe B) pairs with the toehold domain (g*) of probe E and then opens the hairpin structure by displacing the 3′ terminal domain (e and f). The more stable, double-stranded substrate is formed, and the single-stranded region of probe E (g, h, e, f, and i) is released. The e and f domain acts as another toehold for probe F (f*, e*, h*, g*, e, f, g, and h) recognition. Downstream SDR thus occurs in the same manner, opening the hairpin structure of probe F and releasing probe B simultaneously. The displaced probe B triggers more cycles of downstream SDR. Finally, this intelligent design produces many DNA duplexes (probes E and F) on the electrode by trace exosomal miRNA. Since MB is no longer located near the electrode interface, the obtained signal intensity declines sharply, which correlates to the concentration of miRNA input. We have calculated the melting temperature (Tm) of various double-stranded regions during the cascade SDR (see Supporting Information Figure S1). Higher Tm indicates a steadier state, and faster displacement reaction occurs with larger ΔTm values. The calculated values demonstrate that it is thermodynamically favorable from the initial interaction with the target. Scheme 1 | Illustration of the cascade SDR-based biosensor for the analysis of exosomal miRNA. SDR, strand displacement reaction. Download figure Download PowerPoint The feasibility of the cascade SDR is confirmed by gel electrophoresis analysis. For the upstream reaction, target, probe B, and probe C with low molecule weights run faster in the gel, while the triple-stranded substrate formed by probes A, B, and C runs slower, which is due to the larger molecule weight. After further interaction with target, probe C is displaced. Because the target sequence is longer than that of probe C, the newly formed triple-stranded substrate with even larger molecule weight runs slower. In contrast, this triple-stranded DNA can also respond to probe D, which displaces probe B and the miR-21. The formed DNA probes can be easily distinguished in the gel (see Supporting Information Figure S2a). For the downstream reaction, the bands of probes B, E, and F are also shown in the gel image, respectively. The duplex formed by probes B and E runs slower, and the band is much brighter, demonstrating the successful pairing reaction. Consequently, the introduction of probe F displaces probe B and forms a more stable DNA substrate, which is verified by the band that runs even slower (see Supporting Information Figure S2b). These gel results are direct evidence of the upstream and downstream SDR. Electrochemical characterizations are then performed to probe stepwise processes at the electrode interface. EIS results are displayed in Figure 1a. After the modification of probe E, a small hairpin loop appears, reflecting the enlarged interfacial charge-transfer resistance caused by the repulsion between DNA and the electrochemical species. After the incubation with the product of the target-induced upstream reaction, the hairpin loop diameter increases because the localization of probe B on the electrode surface enhances the electronegativity. Next, with added probe F, recycled downstream SDR occurs and the longer double-stranded DNA product contributes to even larger charge-transfer resistance. As shown in the Nyquist diagram (Figure 1a), the diameter of hairpin loop domain increases even more. However, in the absence of miR-21, cascade SDR is not possible, and the electrode interface remains with hairpin-structured probe E, which is demonstrated by the nearly unchanged hairpin domain compared with initial probe E-modified electrode. SWV curves are used to further confirm the conclusion. As depicted in Figure 1b, a significant current peak is observed after probe E immobilization because of the short distance between MB and electrode. TCEP-mediated recycling reduction reaction also enhances the signal. However, after pairing with the upstream SDR product, the hairpin structure of probe E is opened, leading to the increase in the distance between MB and electrode. Current peak thus declines. With downstream SDR, more probe E strands are opened, and more MB molecules are kept inactive. Therefore, the corresponding SWV peak is further decreased. In instances without Target, SDR cannot be triggered, and the peak intensity is barely changed. These results confirm good feasibility of this strategy. Figure 1 | (a) Nyquist diagrams and (b) SWV curves of (A) gold electrode, (B) probe E-modified electrode, (C) after reaction with miRNA-mediated upstream SDR product, (D) after further reaction with probe F, (E) probe E-modified electrode after cascade SDR in the absence of miRNA. Inset in (a) is the electrical equivalent circuit. SWV, square wave voltammetry; SDR, strand displacement reaction. Download figure Download PowerPoint To achieve better analytical performances, some parameters should be optimized. The density of probe E on the electrode surface is important. Larger density contributes to larger signal output but excessively crowded interface may hinder base pairing efficiency. By studying the SWV peak intensity, 0.7 μM is selected for the following experiments (see Supporting Information Figure S3a). Similarly, the concentration of triple-stranded substrate (probes A, B, and C) is optimized at 0.06 μM (see Supporting Information Figure S3b). Under these conditions, different concentrations of target are used to initiate cascade toehold-mediated SDR, and the analytical performances are assessed. In the SWV curves (Figure 2a), the current peak value (ip) decreases when the target concentration rises. The relationship between the change of ip and the logarithm of miR-21 concentration (c) is summarized in Figure 2b. A wide dynamic range from 100 aM to 100 pM is established with the linear regression equation (eq 1): Δ i p = 29.700 + 1.825 Log ( c ) ( n = 4 , R 2 = 0.998 ) (1) Figure 2 | (a) SWV curves for the detection of miRNA with the concentrations of 0, 100 aM, 1 fM, 10 fM, 100 fM, 1 pM, 10 pM, 100 pM, and 10 nM (from top to bottom). (b) Calibration plot representing the relationship between Δip and Log(c). Inset presents the linear range. (c) Regeneration cycles of electrode at different concentrations. (d) Reproducibility of electrode at different concentrations. SWV, square wave voltammetry. Download figure Download PowerPoint The limit of detection is evaluated based on the 3σ rule, which is 40 aM. After comparing with some representative methods reported recently, excellent analytical performance of this method is confirmed (see Supporting Information Table S1). The strategy only involves SDR without polymerization or enzymatic digestion, which allows it to be easily regenerated. We have tested this capability by applying a denature and reanneal cycle of DNA strands on the electrode surface after each test. The current peak intensities are summarized in Figure 2c; reproduced electrode presents similar electrochemical intensity to original probe E immobilized electrode, and the results for the target analysis are quite stable. The reproducibility is then assessed by employing several gold electrodes. These independent electrodes are treated in the same manner for the measurements of different standard samples. The recorded current peak values are quite consistent, and the relative standard deviation (RSD) is <5%, verifying the excellent reproducibility of the proposed work (Figure 2d). To investigate the selectivity of this method, we apply a series of potentially interfering miRNA sequences including miR-29a, miR-155, miR-200c, miR-141, miR-605, and three single-base-mismatched miRNAs (M1, M2, and M3) (see Supporting Information Table S2). SWV curves are measured and analyzed. Compared with miR-21, these miRNAs lead to negligible decrease of ip. However, after adding miR-21, Δip increases to a remarkable level, demonstrating the proposed method is sequence dependent (Figure 3a). We then check its performance in biological samples. Different amounts of the target are spiked in PBS and human serum samples separately, which are then tested. As shown in Figure 3b, peak-current variations of serum samples accord with those of PBS samples, demonstrating excellent utility in biological samples. Figure 3 | (a) Comparison of Δip for miRNAs assay in the presence and absence of target spiking. (b) Practical utility assessments in PBS buffer and serum samples. PBS, phosphate-buffered saline. Download figure Download PowerPoint We then measure exosomal miRNA directly by this strategy. Exosomes are first isolated by ultracentrifugation from cell culture media. As shown in Figure 4a, saucer-like morphology of exosomes is exhibited in the transmission electron microscopy (TEM) image. The diameter distribution of collected exosomes centered at 115 nm is determined by nanoparticle tracking analysis (NTA) (Figure 4b). The size observed in TEM is slightly smaller because of shrinkage during TEM sample preparation. To further verify the identity of exosomes, we conduct western blot experiments. In Figure 4c, the bands demonstrate abundant existence of two exosome-enriched proteins (CD63 and TSG101). These results indicate that we have successfully isolated the exosomes. Next, we employ the proposed method to detect target miR-21 in total RNA extracted from isolated exosomes. Different amounts of exosomes are challenged (Figure 4d). With the increase of exosome numbers, Δip expands for the exosomes from cancer cell lines MCF-7 and MCF 10A. Apparently, the electrochemical response of MCF-7 exosomes is higher, which is due to the overexpression of miR-21 in MCF-7. We then further explore the practical utility of this method by introducing clinical peripheral blood plasma samples, which contain abundant circulating exosomes. The samples are from breast cancer patients and healthy RNA are extracted from these samples for analysis. We SWV peak current variations for patients and healthy the plot (Figure cancer samples with of miR-21 can be easily distinguished from healthy This clinical assay the proposed method may be applied as a powerful for and practical analysis of exosomal miRNA. Figure 4 | (a) TEM and (b) of collected exosomes. (c) blot bands of (d) Comparison of Δip for the analysis of exosomes from different plot shows significant of miR-21 in patients compared with healthy controls transmission electron nanoparticle tracking analysis. Download figure Download PowerPoint We have a novel biosensor for exosomal miRNA analysis based on the design of a cascade toehold-mediated SDR. signal amplification SDR in homogeneous phase and at the solid interface, and TCEP-mediated electrochemical this method ultrahigh selectivity is demonstrated against interfering performance in serum samples with the qRT-PCR technique and has been successfully applied in blood collected from human In addition, electrode and DNA probes can be integrated to form an for practical exosomal miRNA We this work is a good of current liquid and the proposed miRNA biosensor has great potential for early diagnosis of Supporting Information Supporting Information is of is no of to Information This work is by the Science of The to for the 1. by and for of Exosomes from to in a of with Liu Peng Liu Liu on a for Electrochemical Detection of RNA and that and and of Circulating Exosomes Using an Detection of MicroRNA Using of miRNAs and of Exosomes and Exosomal miRNAs Using in and RNA at for and of in and Circulating with Strand for and miRNA Detection at qRT-PCR and MicroRNA in the of miRNA Detection by of a and Detection of in Using Raman on Detection of miRNA with an for of in of and on miRNA from Electrochemical for Detection of for Liu for and Detection on Electrochemical Detection of on Liu Peng of and and Detection of Exosomal Boriachek Electrochemical Detection of Exosomal in Guo of Exosomal MicroRNA on Reaction with Strand Displacement in a for of and Miao Tang Reaction Electrochemical for Ultrasensitive Detection of Exosomal miRNA. and Detection of Exosomal miRNAs for of Using a Raman on Miao Tang of miRNA Strand Displacement and C, Liu and A and DNA for in and of MicroRNA in to the Chemistry, , Information & Chinese wave to for the

Tech Support

Original Source

DOI: https://doi.org/10.31635/ccschem.020.202000458