A novel label free long non-coding RNA electrochemical biosensor based on green L-cysteine electrodeposition and Au–Rh hollow nanospheres as tags

Abstract

Nuclear paraspeckle assembly transcript 1 (NEAT1) that reflects whether the body's anti-HIV virus immunity is strong or weak is one of many long non-coding RNAs (lncRNAs). The effective detection of NEAT1 has great scientific significance and clinical value. However, the detection of lncRNA is difficult with existing methods. The quantification of lncRNA by electrochemical methods based on changes in circuit properties such as capacitance, conductance and resistance may solve this problem. In this study, we describe a new label-free strategy based on catalytic signal amplification with single-wall carbon nanotubes wrapped with Au–Rh hollow nanospheres (Au/Rh-HNP@SWCNT). This strategy is useful for detecting the lncRNA NEAT1. Firstly, L-cysteine (L-Cys) was electrodeposited onto a Au electrode and incubated in colloidal Au to fully expose all of the binding sites of the Au particles, which can bind to L-Cys through its thiol-group. It can take advantage of each surface of the Au nanoparticles that is bound to the capture probe, which contains a (GGG)₃ trimer that can bind to the electron mediator hemin. The results indicate that catalysis was noticeably enhanced and that the biosensor provided ultrasensitive detection of the lncRNA NEAT1. The linear calibration of the biosensor ranged from 1 fM mL⁻¹ to 100 nM mL⁻¹, and the limit of detection was 0.8863 fM mL⁻¹. This lncRNA biosensor based on Au/Rh-HNP@SWCNT complex signal amplification and an L-Cys Au nano-film exhibited acceptable reproducibility and clear selectivity. This strategy may provide a new alternative for clinical HIV diagnosis through the detection of NEAT1.

---

1. Introduction

Long non-coding RNAs (lncRNAs) are nucleic acid transcripts that are >200 nt in length,¹ which affect numerous cellular processes involved in virus replication. Nuclear paraspeckle assembly transcript 1 (NEAT1) is one such lncRNA and is over expressed in patients with human immunodeficiency virus (HIV-1),² which is a serious and fatal worldwide immunodeficiency disease. The anti-virus capacities of patients who over express NEAT1 are much stronger than patients with lower NEAT1 expression. The level of NEAT1 expression reflects whether the body's antiviral ability is strong or weak. Therefore, detection of NEAT1 may be useful in HIV clinical treatment. However, the detection of lncRNA faces several challenges. Detection is typically based on hybridization,³ and traditional molecular biology techniques for lncRNA detection, such as cloning and northern blotting assays, are time-consuming and not very sensitive.⁴ To enhance the sensitivity of lncRNA detection, real-time polymerase chain reaction (rt-PCR) is often used because of exceptional amplification by extending sequences constantly and repeatedly, but professional's personnel and equipment is indispensable.^5,6 So exploiting a rapid, accurate, sensitive and quantitative method for detecting NEAT1 is necessary.

An electrochemical RNA biosensor that uses novel signal amplification and that exhibits improved characteristics may provide a solution for NEAT1 detection. Use of hydrogen peroxide (H₂O₂) in biological systems and practical applications has led to the development of efficient electrochemical H₂O₂ sensors, which are especially interesting to researchers. Various materials such as prussian blue (PB), carbon nanotubes (CNTs) have been used in the construction of H₂O₂ sensors.⁷ Rhodium (Rh) is a noble metal with excellent catalytic activity toward H₂O₂.⁸ It is also useful in hydrogenation reactions^9,10 and oxidation reactions.¹¹ In recent years, the applications of bimetallic nanoparticles for biosensors have attracted particular interest in the fields of electrocatalysis and biosensing. The addition of a second metal in bimetallic catalysts results in the formation of an alloy structure, which, in turn, results in variations in particle size, shape, surface structure, and chemical and physical properties related to the alloy's catalytic activity and chemical selectivity.¹² Janyasupab et al. reported that the addition of Ru into Pt resulted in a catalyst with improved properties compared to the single Pt catalyst for H₂O₂ oxidation detection at a lower potential.¹³ Janyasupab et al. also have described a new strategy involving the use of Pt-based bimetallic nanoparticles containing Cu, Ni, Pd, and Rh for the detection of physiological H₂O₂, which is important in biosensing.¹² Divins et al. have reported a Rh–Pd/CeO₂ bimetallic catalyst nanomaterial. We know that Au nanoparticles can bind to biological macromolecules such as proteins and nucleic acids, through stabilized Au–S bonds.^14,15 If Rh and Au were combined to form bimetallic nanoparticles, the resulting nanoparticles may exhibit the advantages of both metals. In addition to the bifunctional effects and the electronic modification of the two metallic components, the shape can be controlled to improve the catalytic and electro-catalytic properties of the nanomaterials.^16,17 Zhang et al. reported that three-dimensional (3D) porous Pt–Rh nanostructures exhibited enhanced catalytic activity and durability when used in direct methanol fuel cells.¹⁸

L-Cys self-assembled monolayers (SAMs) have been widely used to construct sensors because of their –NH₂ and –SH functional groups.^19,20 L-Cys films have been assembled onto the surface of Au electrodes by electrodeposition, which yields a –NH₂ functional interface for assembling Au nanoparticles. In the presence of L-Cys, the sizes of Au nanoparticles are uniformly distributed in the range of 10–40 nm.²¹ The basic process for SAM formation typically involves immersing the substrate into an L-Cys solution. However, this immersion technique does not prevent possible competitive adsorption by other ions. Consequently, the assemblies formed by immersion growth are sometimes non-uniform. The electrodeposited L-Cys films have several advantages, for example, the electrodeposited films are controllable, compact, stable, and favorable for electron transfer toward the electrode surface.²² Zhuo et al. developed hollow Au nanospheres for the assembly of PAMAM–L-Cys via thiol–Au bonding, which was subsequently used for loading antibodies (Ab₂).²³

Single-wall carbon nanotubes (SWCNTs) offer unique advantages, including enhanced electronic properties, a large edge-plane-to-basal-plane ratio, and rapid electrode kinetics.²⁴ Therefore, SWCNT-based biosensors generally exhibit higher sensitivities, lower detection limits. Many variables must be tested and optimized to create an SWCNT-based sensor.²⁵ Herein, a new SWCNT-encapsulated hollow Au–Rh nanosphere (Au/Rh-HNP@SWCNT) structure was synthesized; the resulting Au/Rh-HNP@SWCNT exhibits the advantage of the biological macromolecule binding features of Au and the superior catalytic properties of Rh. This new nanomaterial is conducive to the rapid combination of biological macromolecules and nanomaterials and to increasing catalytic performance. In additional, to generate current signal, kinds of electroactive indicator, for instance, thionine, methylthionine chloride, were used in electrochemical DNA biosensor. Recently, hemin containing Fe²⁺ is a relatively new electroactive indicator molecule. For hemin can not directly bind to nucleic acid sequence but with G-quadruplex structure,²⁶ it has been widely used as an electro-active indicator in electro-chemical catalysis.^27,28

As such, we propose a new signal strategy for catalytic signal amplification to detect the lncRNA NEAT1. First, L-Cys was electrodeposited onto Au electrode. Then incubated with Au nanoparticles, which bind to L-Cys through its thiol-group and –NH₂. For capture probe (CP) immobilization, traditional method direct binding to electrode surface only make use of one side of the matrix material. Compare with that way, this approach could take advantage of each surface of the Au nanoparticles to bind CP. Herein sequence of CP is an RNA fragment designed to contain a (GGGG) quadruplex that combines with hemin to produce a current signal. A series of results demonstrate that this biosensor is reproducible, highly sensitive, and stable and has optimal selectivity compared to traditional detection methods. This strategy may provide a new alternative for clinical HIV active situation diagnosis through the detection of NEAT1.

2. Experiments

2.1 Reagents

Gold chloride tetrahydrate (HAuCl₄), rhodium(III) chloride hydrate (RhCl₃), poly(vinylpyrrolidone) (PVP), horseradish peroxidase (HRP), NaBH₄, NaOH, L-Cys, poly dimethyl diallyl ammonium chloride (PDDA), hemin, and cyclohexanethiol (HT) were purchased from Sigma-Aldrich Chemical Co. (St. Louis, MO, USA). Phosphate buffered solution (PBS) (pH 7.0) was prepared using 0.01 M Na₂HPO₄ and 0.01 M KH₂PO₄. The prepared solutions were maintained at 4 °C before use. SWCNTs were purchased from Nanjing Xianfeng Nano Co. (Nanjing, China). The buffer for the preparation of the lncRNA probe solutions was Tris–EDTA buffered solution (TE buffer) (10 mM Tris–Cl, pH 7.4, containing 1 mM EDTA).

NEAT1 was used as a representative lncRNA molecule because of its important role in HIV. A sulfhydryl-modified CP (5′-SH-(CH₂)₆-CGCTCGGCCTGGGACGGGGCCCGG-3′) was designed to hybridize to the target RNA near the 5′ end of the sequence. The mock target RNA sequence is 5′-CTGGGCCCCAGGAAG-3′, miRNA-16 (5′-CTCAACTGGTGTCGTGGAGTCGGCAATTCAGTTGAGCGCCAATA-3′), β-actin (5′-TGACGTGGACATCCGCAAAG-3′), gapdh (5′-GATGGGATTTCCATTGATGACA-3′) and CEA were used as interfering substances. All nucleotide segments were synthesized by Sangon Biotech. Co. (Shanghai, China).

2.2 Apparatus

Cyclic voltammetry (CV) and differential pulse stripping voltammetry (DPV) measurements were performed using a CHI 660d electrochemistry workstation (Shanghai CH Instruments, Shanghai, China). The three-compartment electrochemical cell contained a Pt wire auxiliary electrode and a saturated calomel reference electrode (SCE). The working electrode was a gold electrode (GE, diameter 4 mm). Transmission electron microscopy (TEM) measurements were performed on a JEM-1400 microscope (JEOL, Tokyo, Japan). Scanning electron microscopy (SEM) was carried out on an S4800 microscope (Hitachi Co., Japan). XPS analysis was done in Thermo Escalab 250Xi (Massachusetts, USA). Raman spectral analysis was performed on an in Via Raman microscope (Gloucestershire, UK). The pH measurements were performed with a pH meter (MP 230, Mettler-Toledo, Switzerland) and a digital ion analyzer (model PHS-3C, Dazhong Instruments, Shanghai, China).

2.3 Synthesis of Au nanoparticles

The Au nanoparticles were synthesized according to the classic method.^29,30 First, solutions of HAuCl₄ (1%) and sodium citrate (1%) were prepared. Then, 4 mL of HAuCl₄ was added into 44.5 mL of the sodium citrate solution. The resulting solution was heated to boiling under vigorous stirring, and 1.5 mL of sodium citrate was added. Boiling was continued for 30 min, and the solution was subsequently allowed to naturally cool to room temperature. The color of the solution became red, indicating the formation of Au nanoparticles.

2.4 Au/Rh-HNP synthesis

The Au/Rh-HNP was synthesized using a modified version of a previously reported synthesis method. Briefly, CoCl₂·6H₂O (10 mg) and PVP (M_w 40 000, 60 mg) were dissolved in 50 mL of ddH₂O (18.2 M), sonicated for 10 min, and purged with N₂ for 10 min. Then, 1.0 mL of 1% RhCl₃ and 0.2 mL of 1% HAuCl₄ solutions were combined in a separate flask. A freshly prepared solution of NaBH₄ (5 mg in 10 mL of H₂O) was then added dropwise with stirring. After all of the NaBH₄ was added, the resulting mixture was added dropwise to the CoCl₂/PVP mixture with stirring for 20 min. Finally, the sediment was collected by centrifugation and washed several times with H₂O and ethanol. The precipitate was re-suspended in 2 mL of H₂O and stored at 4 °C for use.

2.5 Preparation of CP-adsorbed nanomaterials

Eight milligrams of SWCNTs was added to 16 mL of H₂O followed by the addition of 267 μL of 30% PDDA (M_w < 10 000) to obtain a 0.5 wt% PDDA solution. The solution was ultrasonically stirred to obtain a stable PDDA-SWCNTs suspension (0.5 mg mL⁻¹).

The proposed CP bioconjugate was prepared as follows: the hydrated SWCNTs were mixed with the Au/Rh-HNP solution overnight. After being collected by centrifugation, the Au/Rh-HNP@SWCNT complex was stored at 4 °C. It was subsequently mixed with 20 μL of NEAT1 CP and stirred overnight. The excess CP was removed by repeated rinsing with distilled water. The precipitate was resuspended in 0.5 mL of distilled water. Subsequently, HRP was added to the solution to block any excess binding sites. After the solution was centrifuged at 5000 rpm for 10 min, the supernatant was discarded and the precipitate was resuspended. This step was repeated three times. Finally, the precipitate was resuspended in 1.2 mL of distilled water (Scheme 1A).

Scheme 1 The fabrication scheme of the lncRNA NEAT1 biosensor. GE: gold electrode, CNT: carbon nanotube, Au/Rh–HNP : Au–Rh hollow nanosphere, CP: capture probe, HRP: horseradish peroxidase, HT: hexanethiol.

2.6 Fabrication of the biosensors

Prior to surface modifications, the GE with a diameter of 4 mm was polished repeatedly with a 0.3 μm and 0.05 μm alumina slurries to obtain a mirror-like surface, followed by successive sonication in double-distilled water, anhydrous ethanol and again in double-distilled water, each for 5 min. The GE was then air-dried. Next, an L-Cys film was deposited onto the pretreated GE via potential cycling at 10 mV s⁻¹ from 0 to 1.8 V in a solution containing 20 mM L-Cys and 0.1 M HCl for 40 cycles.²² The GE was then immersed into colloidal Au at room temperature for 3 h to form a monolayer of Au nanoparticles. Afterwards, the resulting electrode was immersed into a NEAT1–CP solution at 4 °C for 8 h. Subsequently, the electrode was soaked in 20 μL of (1 mM) HT at room temperature for 40 min to block nonspecific binding sites. The biosensor was then incubated with the target solution to bind to hemin (Scheme 1B). Above series blocking steps assured that hemin could only bind to target with G-quadruplex. The biosensor was subsequently tested in PBS containing H₂O₂.

2.7 Experimental measurements

Electrochemical experiments were performed in a conventional electrochemical cell containing a three-electrode arrangement. CV measurements of the electrodes were performed in a 5 mM [Fe(CN)₆]^3−/4− solution containing 0.1 M KCl. DPV was performed from −0.6 to 0 V at a sweep rate of 50 mV s⁻¹ in 2 mL of 0.1 M PBS (pH 6.5) with an appropriate amount of H₂O₂ as the substrate.

3. Results and discussion

3.1 TEM and XPS characterization of the Au/Rh-HNP, SWCNT, Au/Rh-HNP@SWCNT complex

As shown in Fig. 1A–C, the nanomaterials were first characterized by TEM. Au/Rh-HNP was synthesized and consisted of numerous hollow Rh spheres dispersed with the monomer. The Au/Rh-HNP′ diameters ranged from 20 to 120 nm. A typical TEM image of a hollow Au–Rh nanosphere clearly shows an empty interior with a black circumference (Fig. 1A), indicating that the hollow structure of the nano-Rh was successfully produced.

Fig. 1 TEM micrographs and of the (A) Au/Rh-HNPs, (B) SWCNTs, (C) Au/Rh-HNP@SWCNT complexes, XPS pattern of (D) Au/Rh-HNP@SWCNT complexes.

The morphology of PDDA-hydrated SWCNTs is shown in Fig. 1B. The SWCNTs often exhibit bent long lines with a hollow structure. The SWCNTs are made from graphite with curled structures. The diameter of the SWCNTs typically ranged between 5 and 6 nm, and their length ranged from 400 nm to 2 μm. Because of the hydration effect of PDDA, some small pieces of mist were observed beside the SWCNTs and generally overlapped each other.

After continual agitation of the Au/Rh-HNPs and SWCNTs for several hours, Au/Rh-HNP@SWCNT was synthesized. The TEM results demonstrate that the Au/Rh-HNP visibly combines with SWCNT and forms a Au/Rh-HNP@SWCNT complex (Fig. 1C). The results clearly show that Au/Rh-HNP@SWCNT complexes were synthesized.

To analyze whether the components of the Au/Rh-HNP@SWCNT complex were complete and met our expectations, we conducted XPS analysis. The results show that the complex contains four elements: C, O, Au, and Rh (Fig. 1D). The C : O ratio is approximately 1 : 1, which is appropriate for SWCNTs. The Rh : Au ratio is approximately 5.8 : 1, consistent with the expected ratio for the Au/Rh-HNP@SWCNT complex.

3.2 SEM analysis of the Au/Rh-HNPs, SWCNTs, and Au/Rh-HNP@SWCNT complexes

These nanomaterials were further characterized by SEM for a more comprehensive understanding of their surface structure. Individual Au/Rh-HNPs exhibit distinct differences in brightness between the inner and outer regions of the particles. The average diameter of the Au/Rh-HNPs was estimated to be 20–140 nm (Fig. 2A), which is consistent with the TEM results. The Au/Rh-HNPs exhibited a substantially greater specific surface area for the catalytic reaction compared to the solid Rh or Au nanoparticles, which resulted insubstantially higher binding and catalytic capacities.

Fig. 2 SEM micrographs of the (A) Au/Rh-HNPs, (B) SWCNTs, (C) Au/Rh-HNP@SWCNT complexes.

As shown in Fig. 2B, the SWCNT hydrated by PDDA alone exhibits numerous cluster and crumb lines with a bent or straight shape. Numerous small molecules are observed in a ball around the SWCNTs; these objects are believed to be PDDA combined with SWCNTs and to correspond to the water mist observed in the TEM micrographs.

The SEM results showed that the Au/Rh-HNPs and SWCNTs form a complex in which Au/Rh-HNPs clearly bind to SWCNTs (Fig. 2C). These results are supported by the previously discussed TEM images.

3.3 Raman spectral analysis of Au/Rh-HNP@SWCNT complexes

Raman spectroscopy is a powerful tool for characterizing the structure of complexes such as Au/Rh-HNP@SWCNT. Raman spectra were obtained at an excitation wave length of λ = 532 nm; the corresponding results are shown in Fig. 3. The samples exhibit three prominent peaks at approximately 1347 cm⁻¹, 1590 cm⁻¹ and 2692 cm⁻¹ (Fig. 3). The peaks at 1347 cm⁻¹, 1590 cm⁻¹ and 2692 cm⁻¹ are attributed to the D, G and G′ bands of the SWCNTs, respectively. Previously reported results^31,32 support the results of our analysis. The D band is usually related to the breathing mode of the k-point phonons of A_1g symmetry, whereas the G band is associated with the E_2g vibration mode of sp² carbon domains. The effect of boron and nitrogen doping appears in the high- and low-energy sides of the original component in G′ band spectra, and the energy of the new component is closely related to p- or n-type doping.³³ When compared with the previously reported Raman D, G and G′ bands for pure SWCNTs,³¹ the D, G and G′ bands are shifted +13, +4 and +32 cm⁻¹, respectively. These changes are attributed to effects of the Au/Rh-HNPs.

Fig. 3 Raman spectrum of the Au/Rh-HNP@SWCNT complex.

3.4 Electrodeposited L-Cys/Au nanoparticle reaction platform for the proposed biosensor

The L-Cys films were electrodeposited onto a pretreated GE by potential cycling at 10 mV s⁻¹ from of 0 to 1.8 V for 40 cycles in a solution containing 20 mM L-Cys and 0.1 M HCl. The L-Cys was bound to the GE via an electric current that resulted in the formation of a Au–S bond. As shown in Fig. 4A, a typical cyclic voltammogram for the electrolysis includes two anodic peaks and two cathodic peaks (Fig. 4A). As the number of cycles increases, the anodic peak current gradually decreases and the L-Cys film gradually forms on the electrode, as indicated by the arrow.

Fig. 4 (A) CVs for 40 cycles of electrolysis in 1.0 mg mL⁻¹ Cys in pH 7.4 PBS at a scan rate of 50 mV s⁻¹. (B) CV changes during L-Cys electrolysis: (a) before electrolysis; (b) after electrolysis. (C) SEM micrograph of Au nanoparticles.

The deposition effects were further confirmed by electrochemical characterization in a K₃[Fe(CN)₆] solution (5.0 mM) at a scan rate of 50 mV s⁻¹. In Fig. 4B, the blue curve (a) represents the result for the bare GE, and the red curve (b) represents the result for the electrode after electrodeposition of the L-Cys film. As shown, the redox current peak of curve (b) is clearly lower than that of curve (a) (ΔI = 104.07 μA), indicating a highly efficient electrodeposition of L-Cys. The reason for this high efficiency is that the L-Cys blocks electron transfer by electrodeposition. The films can immobilize Au particles (Fig. 4C) through Au–S bonding. The distribution of Au particles is relatively uniform, and the film thickness can be controlled by the number of cycles.

3.5 CV characteristics of the NEAT1 biosensor

The CV results support a stepwise insulation of the conductive surface as the electrode is modified. Electrochemical characterizations of differently modified electrodes were tested in K₃[Fe(CN)₆] solution (5.0 mM) with a CV method at a scan rate of 50 mV s⁻¹. As shown in Fig. 5, curve (a) represents the bare GE, and curve (b) represents the L-Cys film. Curve (a) shows a pair of typical reversible redox peaks that correspond to the reversible redox reaction of the bare GE in a ferricyanide solution. The redox peak currents of curve (b) remarkably decreased when compared to those of curve (a), indicating that the L-Cys film was successfully electrodeposited onto the electrode. The electrode was also incubated with the Au nanoparticles, as represented by curve (c). Curve (c) was notably higher than curve (b), implying that the Au nanoparticles were abundantly immobilized on the L-Cys film and could promote the transmission of electrons to the electrode surface. A comparison of the Cys/nano Au/CP (curve (d)) with the Cys/nano Au (curve (c)) reveals that the peak currents of curve (d) were increased because of the high conductivity of the Au/Rh-HNP@SWCNT–CP complex. Curve (e) represents Cys/nano Au/CP/HT, and the peak currents of curve (e) were significantly lower than those of curve (d). Curve (f) represents Cys/nano Au/CP/HT/NEAT1, which exhibited a slightly lower peak current for the limited small amount of target NEAT1 sequence. We concluded that a series of RNAs, including CP, NEAT1 and HT, all abundantly bind to the electrode during CV characterization.

Fig. 5 Cyclic voltammograms of different modified electrodes measured in a 5 mM [Fe(CN)₆]^3−/4− solution containing 0.1 M KCl (pH 7.4): (a) bare GE; (b) deposition of Cys; (c) Cys/nano Au; (d) Cys/nano Au/CP; (e) Cys/nano Au/CP/HT; (f) Cys/nano Au/CP/HT/NEAT1.

3.6 Catalytic performance of the biosensor

The signal amplification of our biosensor was based on Au/Rh-HNP–HRP. To analyze the catalytic effect of Au/Rh-HNP–HRP on H₂O₂, H₂O₂ was added to 2 mL of PBS (pH 7.0), and the response of the biosensor was measured using DPV. The results showed that the current peak was amplified upon the addition of H₂O₂ because of the high catalytic activity of Au/Rh-HNP–HRP and that the current peak clearly increased after the H₂O₂ was added. The current response was amplified approximately 4-fold (Fig. 6), indicating a catalytic effect that substantially exceeds those previously reported.^12,34 Therefore, the high sensitivity is a result of multiple signal amplifications, indicating that the proposed strategy is appropriate for the detection of low concentrations of H₂O₂ in small sample volumes.

Fig. 6 The catalytic activity in 2 mL of PBS (pH 7.0): (a) without H₂O₂; (b) with H₂O₂.

3.7 Optimization of experimental parameters

The optimized pH of the reaction solution was determined. A fully assembled biosensor coating consisting of 1 pM mL⁻¹ of NEAT1 was dipped into 2 mL of 0.1 M PBS containing H₂O₂ and tested by DPV methods. As shown in Fig. 7A, the pH of the working buffer was measured by adding the proposed biosensor to the working buffer solutions with pH values ranging from 5.0 to 8.0. The results indicate that the peak current response was into plateau after pH reach 6.5 to 7.5. Thus, PBS with a pH of 6.5 was determined to be the optimum condition and was used as the working buffer in subsequent studies.

Fig. 7 Optimization of (A) the pH of the CP–NEAT1 reaction and (B) the concentration of H₂O₂ between 0 and 1.8 mM L⁻¹.

In addition, the concentration of H₂O₂ in the detection solution was tested by DPV for various concentrations of H₂O₂; this detection was based on the biosensor reaction with 1 pM NEAT1 (Fig. 7B). The results indicated that the current responses increased as the concentration of H₂O₂ was increased to 1.2 mM and began to plateau at higher concentrations. At 1.4 mM H₂O₂, the current was inhibited by excess H₂O₂. Therefore, 1.2 mM H₂O₂ was used as the optimum concentration for further studies.

3.8 Calibration of the biosensor for NEAT1 detection

The proposed electrochemical biosensor was tested by incubation with various NEAT1 concentrations. As shown in the inset of Fig. 8, the cathodic peak clearly decreased in intensity with increasing NEAT1 concentration, which we attributed to the bulky NEAT1 attached to the electrode surface. NEAT1 created a barrier for the electrons and substantially inhibited electron transfer. Fig. 8 displays the corresponding calibration plots. The cathodic peak currents were proportional to the NEAT1 concentration over a concentration range from 1 fM mL⁻¹ to 100 nM mL⁻¹. A linear relationship was observed in the concentration range from 0 to 100 nM mL⁻¹, and the linear equation was I = 3.503 log C_NEAT1 − 24.170; the correlation coefficient was 0.9921, and the detection limit was 0.8863 fM mL⁻¹ (defined as 3σ/κ, where σ is the standard deviation of the blank and κ is the slope of the corresponding calibration curve). The detection limit was much lower than those obtained using traditional methods.

Fig. 8 Calibration plots of the cathodic peak current response vs. the NEAT1 concentration. The inset shows the CVs of the cathodic peak at various concentrations (from top to bottom: 0, 1.00 × 10⁻³, 5.00 × 10⁻³, 1.00 × 10⁻², 1.00 × 10⁻¹, 1.00, 1.00 × 10¹, 1.00 × 10², 1.00 × 10³, 1.00 × 10⁴ and 1.00 × 10⁵ pM mL⁻¹).

3.9 Reproducibility, repeatability, specificity and stability of the proposed lncRNA biosensor

The reproducibility was evaluated over 50 continuous cycles and over a potential range of −0.6 to 0 V at a scan rate of 50 mV s⁻¹ in 0.1 M PBS (Fig. 9A). The redox lines of all 50 cycles precisely coincided, and the relative standard deviation was less than 1.9%, indicating that the biosensor was perfectly stable when used to scan a buffer solution.

Fig. 9 (A) Results of 40 CV cycles in 2 mL of PBS (pH 6.5) after the electrodes was incubated with NEAT1. (B) CVs of the modified electrode at different scan rates (from inner to outer): 10, 50, 100, 200, 300 mV s⁻¹ in 2 mL of PBS (pH 6.5) at room temperature.

Fig. 9B shows the CVs of the modified biosensor in 0.1 M PBS at scan rates ranging from 10 to 300 mV s⁻¹ (i.e., at 10, 50, 100, 200 and 300 mV s⁻¹). The anodic and cathodic peaks clearly varied with the scan rate. Moreover, the peak currents were directly proportional to the scan rates (Fig. 9B). These results reveal a surface-confined redox process on the electrode.

To confirm whether the observed voltammetric response was generated by a CP–NEAT1 target-specific interaction or by a non-specific nucleotide interaction, we tested the selectivity when the RNA biosensor was incubated with samples containing the following potential interfering substances: gapdh, β-actin, AFP, miRNA16, NEAT1 and NEAT1 mixture. Even in the presence of a 50-fold excess of RNA (50 pM, gapdh, β-actin, miRNA16) and protein (5 ng CEA), minimal current responses were detected, similar to the results of the blank test. However, the presence of the perfectly matched target lncRNA NEAT1 and its mixture at an even lower concentration (50-fold, 1 pM) resulted in a significant increase in the current response (Fig. 10). This gap is statistically significant according to calculations performed using SPSS 19.0 (P < 0.05). These results reveal that the lncRNA biosensor exhibits ideal specificity.

Fig. 10 Specificity of the NEAT1 biosensor with 50 pM of gapdh, 50 pM of actin, 5 ng of CEA, 50 pM of miRNA16, 1 pM of NEAT1 and a mixture (1 pM NEAT1, 50 pM of gapdh, 50 pM of actin and 5 ng of CEA).

The stability of the NEAT1 biosensor was evaluated for 33 days of storage at 4 °C and was measured every 3 days. The DPV peak current of the RNA biosensor decreased gradually, and the final peak current retained 96.27% of the initial current after 27 days, demonstrating the acceptable stability of the lncRNA biosensor.

3.10 Synthetic serum sample analysis

Recovery experiments were performed to monitor the feasibility of the prepared biosensor by standard addition methods.^27,35 Gradient concentrations of the synthetic NEAT1 segments (0.5, 1, 5, 10, 20, 80 pM mL⁻¹) were added into serum and tested using the studied methods. As shown in Table 1, some results of the synthetic serum sample test obtained using the proposed biosensor were acceptable, with recoveries between 97.8% and 103.0% (Table 1); these results indicate that the proposed biosensor is feasible and can satisfy the requirements for practical analyses.

Table 1 Recovery experiments of lncRNA NEAT1 in serum samples

Serum sample	Add (pM)	Response (μA)	Found (nM)	Recovery (%)
1	0.5	−25.26	0.489	97.8
2	1	−24.20	0.981	98.1
3	5	−21.68	5.14	102.8
4	10	−20.69	9.85	98.5
5	20	−19.57	20.59	103.0
6	80	−17.48	81.03	101.3

---

4. Conclusions

In conclusion, we have developed a novel biosensor using Au/Rh-HNP@SWCNT complex and a reaction platform of an L-Cys film–Au nanoparticle to detect NEAT1. The L-Cys film forms through rapid electrodeposition to bind to Au nanoparticles at exposed binding sites. In addition, the Au/Rh-HNP@SWCNTs was used to label the CP of NEAT1 as tags. The Au particles may promote greater binding speed between the CP and hollow Rh spheres, which have a catalytic function. Moreover, we utilized cooperative signal amplification strategies based on the high catalytic activity of Rh-HNPs and HRP for the reduction of H₂O₂. The NEAT1 biosensor exhibited high sensitivity, a low detection limit, acceptable reproducibility, and desirable specificity. This biosensor may prove to be useful in monitoring HIV curative effects through the detection of NEAT1.

Abbreviations

lncRNAs	Long non-coding RNAs
NEAT1	Nuclear paraspeckle assembly transcript 1
HIV	Human immunodeficiency virus
rt-PCR	Real-time polymerase chain reaction
SAMs	Self-assembled monolayers
L-Cys	L-Cysteine
CNT	Carbon nanotube
PB	Prussian blue
SWCNTs	Single-wall carbon nanotubes
Au/Rh-HNP	Au–Rh nanosphere
PVP	Poly(vinylpyrrolidone)
HRP	Horseradish peroxidase
HT	Cyclohexanethiol
CV	Cyclic voltammetry
DPV	Differential pulse stripping voltammetry
GE	Gold electrode
TEM	Transmission electron microscopy
SEM	Scanning electron microscopy
CP	Capture probe
XPS	X-ray photoelectron spectroscopy
PDDA	Poly dimethyl diallyl ammonium chloride
PBS	Phosphate buffered solution

Acknowledgements

This work was supported by the National Natural Science Foundation of China (NSFC 81401753, NSFC 81371898). We appreciate the valuable comments that we received from other members of our laboratories.

References

1. Q. Zhang and K. T. Jeang, Biomed. Pharmacother., 2013, 3, 34–42.
2. Q. Zhang, C. Y. Chen, V. S. Yedavalli and K. T. Jeang, mBio, 2013, 4, e00596–e00605.
3. K. A. Cissell and S. K. Deo, Anal. Bioanal. Chem., 2009, 394, 1109–1116.
4. W. Kong, L. L. He, M. Coppola, J. P. Guo, N. N. Esposito, D. Coppola and J. Q. Cheng, J. Biol. Chem., 2010, 285, 17869–17879.
5. M. L. Wong and J. F. Medrano, BioTechniques, 2005, 39, 75.
6. S. A. Bustin, V. Benes, T. Nolan and M. W. Pfaffl, J. Mol. Endocrinol., 2005, 34, 597–601.
7. W. Chen, S. Cai, Q. Q. Ren, W. Wen and Y. D. Zhao, Analyst, 2012, 137, 49–58.
8. H. Zhu, M. Zhang, S. Y. Cai, Y. T. Cai, P. Wang, S. Y. Bao, M. L. Zou and M. L. Du, RSC Adv., 2014, 4, 794–804.
9. I. S. Park, M. S. Kwon, N. Kim, J. S. Lee, K. Y. Kang and J. Park, Chem. Commun., 2005, 5667–5669.
10. I. S. Park, M. S. Kwon, K. Y. Kang, J. S. Lee and J. Park, Adv. Synth. Catal., 2007, 349, 2039–2047.
11. M. E. Grass, Y. Zhang, D. R. Butcher, J. Y. Park, Y. Li, H. Bluhm, K. M. Bratlie, T. Zhang and G. A. Somorjai, Angew. Chem., Int. Ed., 2008, 47, 8893–8896.
12. M. Janyasupab, C. W. Liu, Y. Zhang, K. W. Wang and C. C. Liu, Sens. Actuators, B, 2013, 179, 209–214.
13. M. Janyasupab, Y. Zhang, P.-Y. Lin, B. Bartling, J. Xu and C.-C. Liu, J. Nanotechnol., 2011, 2011, 6.
14. J. M. Han, J. Ma and Z. F. Ma, Electrochem. Commun., 2013, 33, 47–50.
15. J. M. Han, J. Ma and Z. F. Ma, Biosens. Bioelectron., 2013, 47, 243–247.
16. L. Zhang, J. W. Zhang, Q. Kuang, S. F. Xie, Z. Y. Jiang, Z. X. Xie and L. S. Zheng, J. Am. Chem. Soc., 2011, 133, 17114–17117.
17. K. B. Zhou and Y. D. Li, Angew. Chem., Int. Ed., 2012, 51, 602–613.
18. Y. Zhang, M. Janyasupab, C.-W. Liu, X. Li, J. Xu and C.-C. Liu, Adv. Funct. Mater., 2012, 22, 3570–3575.
19. M. Milovanović, A. Djeković, V. Volarević, B. Petrović, N. Arsenijević and Ž. D. Bugarčić, J. Inorg. Biochem., 2010, 104, 944–949.
20. G. Jie, B. Liu, H. Pan, J.-J. Zhu and H.-Y. Chen, Anal. Chem., 2007, 79, 5574–5581.
21. A. Dolati, I. Imanieh, F. Salehi and M. Farahani, Mater. Sci. Eng., B, 2011, 176, 1307–1312.
22. H. A. Niu, R. Yuan, Y. Q. Chai, L. Mao, Y. L. Yuan, Y. Zhuo, S. R. Yuan and X. Yang, Biosens. Bioelectron., 2011, 26, 3175–3180.
23. Y. Zhuo, G. Gui, Y. Chai, N. Liao, K. Xiao and R. Yuan, Biosens. Bioelectron., 2014, 53, 459–464.
24. S. Daniel, T. P. Rao, K. S. Rao, S. U. Rani, G. R. K. Naidu, H. Y. Lee and T. Kawai, Sens. Actuators, B, 2007, 122, 672–682.
25. C. B. Jacobs, M. J. Peairs and B. J. Venton, Anal. Chim. Acta, 2010, 662, 105–127.
26. D. Sen and L. C. H. Poon, Crit. Rev. Biochem. Mol. Biol., 2011, 46, 478–492.
27. Y. Li, J. Deng, L. C. Fang, K. K. Yu, H. Huang, L. L. Jiang, W. B. Liang and J. S. Zheng, Biosens. Bioelectron., 2015, 63, 1–6.
28. Y. Li, X. Li, X. Ji and X. Li, Biosens. Bioelectron., 2011, 26, 4095–4098.
29. G. Frens, Nature Phys. Sci., 1973, 241, 20–22.
30. Y. C. Yeh, B. Creran and V. M. Rotello, Nanoscale, 2012, 4, 1871–1880.
31. S. Suzuki and H. Hibino, Carbon, 2011, 49, 2264–2272.
32. S. Costa, E. Borowiak-Palen, M. Kruszynska, A. Bachmatiuk and R. J. Kalenczuk, Mater. Sci.–Pol., 2008, 26, 433–441.
33. I. O. Maciel, N. Anderson, M. A. Pimenta, A. Hartschuh, H. Qian, M. Terrones, H. Terrones, J. Campos-Delgado, A. M. Rao and L. Novotny, Nat. Mater., 2008, 7, 878–883.
34. K. S. Lokesh, Y. Shivaraj, B. P. Dayananda and S. Chandra, Bioelectrochemistry, 2009, 75, 104–109.
35. S. L. Zhang, L. Zhang, X. Zhang, P. H. Yang and J. Y. Cai, Analyst, 2014, 139, 3629–3635.

---