Genosensing Applications of Glassy Carbon Electrodes Modified with Multi-Walled Carbon Nanotubes Non-Covalently Functionalized with Polyarginine

We report the advantages of glassy carbon electrodes (GCE) modified with multi-walled carbon nanotubes (MWCNTs) non-covalently functionalized with polyarginine (PolyArg) for the adsorption and electrooxidation of different DNAs and the analytical applications of the resulting platform. The presence of the carbon nanostructures, and mainly the charge of the PolyArg that supports them, facilitates the adsorption of calf-thymus and salmon sperm double-stranded DNAs and produces an important decrease in the overvoltages for the oxidation of guanine and adenine residues and a significant enhancement in the associated currents. As a proof-of-concept of possible GCE/MWCNTs-PolyArg biosensing applications, we develop an impedimetric genosensor for the quantification of microRNA-21 at femtomolar levels, using GCE/MWCNTs-PolyArg as a platform for immobilizing the DNA probe, with a detection limit of 3fM, a sensitivity of 1.544 × 103 Ω M−1, and a successful application in enriched biological fluids.


Introduction
Electrochemical (bio)sensors have emerged as important analytical tools for the quantification of analytes of clinical, pharmaceutical, environmental, and foods safety relevance [1][2][3], and currently there is an increasing interest in the development of new biosensing platforms and new transduction schemes. The incorporation of carbon nanomaterials has largely improved the analytical performance of the resulting bioanalytical platforms because of their unique physical and chemical properties [4][5][6][7][8]. Among these carbon nanostructures, carbon nanotubes (CNTs) have been widely used for the construction of electrochemical sensors [9][10][11].
The goal of this work is to evaluate the adsorption and electrooxidation of different DNAs at glassy carbon electrodes (GCE) modified with MWCNTs non-covalently functionalized with PolyArg and discuss, as a proof-of-concept, the analytical usefulness of the resulting GCE/MWCNTs-PolyArg to build an impedimetric genosensor for miRNA-21 (miRNA-21) through the use of MWCNTS-PolyArg as a platform for anchoring the DNA probe. DNA-based nanostructured sensors combine the advantages of highly advanced nanomaterials and DNA receptors in one architecture. In the following sections, we discuss the electrochemical response of salmon sperm and calf-thymus double-stranded DNAs at GCE/MWCNTs-PolyArg as well as the optimization of the miRNA-21 biorecognition layer and the analytical performance of the resulting genosensor.

Reagents and Materials
Double-stranded calf-thymus DNA (dsDNA CT , catalog number D-4522), salmon sperm double-stranded DNA (dsDNA S ), poly-L-arginine (PolyArg, mol wt > 70,000), hydroquinone (HQ), and benzoquinone (Bz) were purchased from Sigma-Aldrich (USA). Multi-walled carbon nanotubes powder (MWCNTs) (30 ± 15 nm diameter, 1-5 microns length) was obtained from NanoLab (USA). Ethanol was purchased from J. T. Baker. Other chemicals were reagent grade and used without further purification. The oligonucleotides were purchased from Life Technology (USA) and the sequences are the following: Ultrapure water (ρ = 18.2 MΩ cm) from a Millipore Milli-Q system was used for pre paring all the solutions.

Apparatus
Electrochemical measurements were performed with an Epsilon (BAS) potentiostat. Glassy carbon electrodes (GCE, 3 mm diameter, CHI), either bare or modified with MWC-NTs functionalized with PolyArg, were used as working electrodes. As counter and reference electrodes, we used a platinum wire and Ag/AgCl, 3M NaCl, respectively. All potentials refer to the latter.
The sonication of the MWCNTs-PolyArg mixture was conducted with an ultrasonic processor (VCX 130W, Sonics and Materials, Inc., Newtown, CT, USA), of 20 kHz frequency with a titanium alloy microtip. An Allegra TM 21 ultracentrifuge (Beckman Coulter) with a F2402H rotor was used to centrifuge the samples after sonication.
Electrochemical impedance spectroscopy (EIS) experiments were performed in a 1.0 × 10 −3 M HQ/Bz solution (prepared in 0.050 M phosphate buffer solution pH 7.40) by applying a perturbation of sinusoidal potential of 10 mV in the frequency range of 10 5 -10 −1 Hz and a working potential of 0.200V. The spectra were analyzed and fitted using the Z-view program.
The spectrophotometric experiments were performed with a Shimadzu UV 1700 Pharma spectrometer.

Preparation of MWCNTs Functionalized with PolyArg (MWCNTs-PolyArg)
The dispersion of MWCNTs with PolyArg was prepared by sonicating 0.75 mg of MWCNTs with 1.0 mL of PolyArg solution (1.0 mg mL −1 , prepared in water) with a sonicator probe (amplitude 50%) for 5.0 min. The amplitude was 50% and the mixture was kept in an ice bath during the procedure. After this treatment, the samples were centrifuged at 1500 rpm for 15 min and the supernatant collected for further work (See scheme shown in Figure 1). kept in an ice bath during the procedure. After this treatment, the samples were cen fuged at 1500 rpm for 15 min and the supernatant collected for further work. (See sche shown in Figure 1).

Preparation of GCE Modified with MWCNT-PolyArg (GCE/MWCNT-PolyArg)
GCEs were first polished with alumina slurries of 1.0, 0.3, and 0.05 μm for 1.0 m each, rinsed carefully with water, sonicated for 10 s in water/ethanol 50:50 v/v, and dr under a nitrogen stream. GCE/MWCNT-PolyArg was obtained by dropping 10 μL MWCNT-PolyArg onto the glassy carbon surface and dried at room temperature.

Preparation of GCE/MWCNTs-PolyArg Modified with DNA
The DNAs were adsorbed by dropping at GCE or GCE/MWCNT-PolyArg 10 μL the corresponding solution prepared in acetate buffer solution (0.200 M, pH 5.00). On the solvent was evaporated, the electrodes were washed by immersion in the acet buffer solution for 5 s and then transferred to the stripping solution. The determinat was performed by Adsorptive Stripping Linear Scan Voltammetry (AdSLSV) in a 0.200 acetate buffer solution pH 5.00 at 0.050 V s −1 . The analytical signals were obtained af subtracting the background currents.

microRNA-21 Genosensor
The DNA probe (oligo21, DNAp) was immobilized at GCE/MWCNT-PolyArg acco ing to the protocol previously indicated in Section 2.5. The electrode was washed by i

Adsorption and Electrooxidation of Different DNAs at GCE/MWCNT-PolyArg
Figure 3A displays linear scan voltammograms (LSV) obtained in a 0.200 M acetate buffer solution pH 5.00 before (a,c) and after (b,d) the deposition of 100 ppm dsDNAs at bare GCE (a,b) and GCE/MWCNTs-PolyArg (c,d). At GCE, there are two oxidation peaks, at 0.86 V and 1.31 V, due to the oxidation of guanine and adenine residues [28,29]. The LSV obtained at GCE/MWCNTs-PolyArg shows two oxidation peaks at potentials 93 and 223 mV more negative with currents 32.5 and 26.9 times higher than those obtained at bare GCE, for guanine and adenine, respectively. This decrease in the overvoltages and the enhancement in the oxidation currents are clear evidence of the facilitated interaction of the nucleic acid with the walls of the MWCNTs and the positively charged polycation that wraps them. It is important to mention that the oxidation of 5.0 × 10 −4 M guanine and adenine at GCE/MWCNTs-PolyArg occurs at 0.64 V and 1.07 V, respectively (not shown). Figure 3B depicts the voltammetric response obtained in a 0.200 M acetate buffer solution pH 5.00 before (a,c) and after (b,d) the deposition of 100ppm dsDNACT at bare GCE (a,b) and at GCE/MWCNTs-PolyArg (c,d). In the case of this polynucleotide, the presence of MWCNTs-PolyArg produced a very important decrease in the overvoltages for the oxidation of guanine (Ep = 0.640 V versus 1.100 V for GCE/MWCNTs-PolyArg and GCE, respectively) and adenine (Ep = 1.120 V versus 1.300 V for GCE/MWCNTs-PolyArg and GCE, respectively), due to the considerably higher number of base-pairs of dsDNACT and,   [28,29]. The LSV obtained at GCE/MWCNTs-PolyArg shows two oxidation peaks at potentials 93 and 223 mV more negative with currents 32.5 and 26.9 times higher than those obtained at bare GCE, for guanine and adenine, respectively. This decrease in the overvoltages and the enhancement in the oxidation currents are clear evidence of the facilitated interaction of the nucleic acid with the walls of the MWCNTs and the positively charged polycation that wraps them. It is important to mention that the oxidation of 5.0 × 10 −4 M guanine and adenine at GCE/MWCNTs-PolyArg occurs at 0.64 V and 1.07 V, respectively (not shown). Figure 3B depicts the voltammetric response obtained in a 0.200 M acetate buffer solution pH 5.00 before (a,c) and after (b,d) the deposition of 100ppm dsDNA CT at bare GCE (a,b) and at GCE/MWCNTs-PolyArg (c,d). In the case of this polynucleotide, the presence of MWCNTs-PolyArg produced a very important decrease in the overvoltages for the oxidation of guanine (E p = 0.640 V versus 1.100 V for GCE/MWCNTs-PolyArg and GCE, respectively) and adenine (E p = 1.120 V versus 1.300 V for GCE/MWCNTs-PolyArg and GCE, respectively), due to the considerably higher number of base-pairs of dsDNA CT and, consequently, the higher negative charge density of the sugar-phosphate backbone. Compared to GCE, there is a significant enhancement of the associated currents at a factor of 16.7 and 6.5 for guanine and adenine, respectively, although it is smaller than the one observed for dsDNA S . This less pronounced enhancement in the currents is due to the larger size of dsDNA CT that makes the accessibility for the electrooxidation of guanine and adenine residues more difficult. In fact, the currents are the result of a compromise between the accessibility of the guanine/adenine sites to the carbon surface, the surface packing, and the orientation of the polynucleotide [30][31][32]. The results discussed above demonstrate the relevance of the PolyArg that supports the MWCNTs, not only to disperse them but also to make more favorable the interaction of DNAs at the electrode surface.

Adsorption and Electrooxidation of Different DNAs at GCE/MWCNT-PolyArg
consequently, the higher negative charge density of the sugar-phosphate backbone. Compared to GCE, there is a significant enhancement of the associated currents at a factor of 16.7 and 6.5 for guanine and adenine, respectively, although it is smaller than the one observed for dsDNAS. This less pronounced enhancement in the currents is due to the larger size of dsDNACT that makes the accessibility for the electrooxidation of guanine and adenine residues more difficult. In fact, the currents are the result of a compromise between the accessibility of the guanine/adenine sites to the carbon surface, the surface packing, and the orientation of the polynucleotide [30][31][32]. The results discussed above demonstrate the relevance of the PolyArg that supports the MWCNTs, not only to disperse them but also to make more favorable the interaction of DNAs at the electrode surface.

GCE/MWCNT-PolyArg as a Platform to Build a miRNA-21 Genosensor
As proof-of-concept of the biosensing usefulness of MWCNTs-PolyArg nanohybrid, we develop an impedimetric biosensor for the quantification of miRNA-21. The biorecognition layer of the genosensor was obtained by adsorption of oligo21 at GCE/MWCNTs-PolyArg as probe. Figure 4 shows the effect of the adsorption time of oligo21 at GCE/MWCNTs-PolyArg on the guanine oxidation current. As the adsorption time increases, there is an increment in the oxidation signals, which starts to level off at 100 min, indicating a saturation of the available sites for the adsorption of the nucleic acid. For

GCE/MWCNT-PolyArg as a Platform to Build a miRNA-21 Genosensor
As proof-of-concept of the biosensing usefulness of MWCNTs-PolyArg nanohybrid, we develop an impedimetric biosensor for the quantification of miRNA-21. The biorecognition layer of the genosensor was obtained by adsorption of oligo 21 at GCE/MWCNTs-PolyArg as probe. Figure 4 shows the effect of the adsorption time of oligo 21 at GCE/MWCNTs-PolyArg on the guanine oxidation current. As the adsorption time increases, there is an increment in the oxidation signals, which starts to level off at 100 min, indicating a saturation of the available sites for the adsorption of the nucleic acid. For comparison, the guanine oxidation signal obtained after deposition of 100 ppm oligo 21 at GCE/MWCNTs-PolyArg and solvent evaporation until dryness is also included. This current shows no significant difference with the one obtained after 2 h adsorption. Figure 4B displays the guanine oxidation currents for increasing concentrations of oligo 21 deposited at GCE/MWCNTs-PolyArg. The currents almost level off for concentrations higher than 50 ppm. Therefore, the selected conditions to ensure the full coverage of the surface were the deposition of an aliquot of 100 ppm oligo 21 solution at GCE/MWCNTs-PolyArg surface followed by the evaporation of the solvent. GCE/MWCNTs-PolyArg and solvent evaporation until dryness is also include rent shows no significant difference with the one obtained after 2 h adsorption displays the guanine oxidation currents for increasing concentrations of oligo2 at GCE/MWCNTs-PolyArg. The currents almost level off for concentrations h 50 ppm. Therefore, the selected conditions to ensure the full coverage of the su the deposition of an aliquot of 100 ppm oligo21 solution at GCE/MWCNTs-Po face followed by the evaporation of the solvent.     The calibration plot presented a linear relationship between 10 −14 M and 10 −12 M miRNA-21, with a sensitivity of 1.544 × 10 3 Ω M −1 (r 2 = 0.992), a detection limit (DL) of 3.3 fM, and a quantification limit (QL) of 10fM (calculated according to 3σ/S and 10σ/S, for DL and QL, respectively, where S is the sensitivity and σ is the standard deviation of the blank signal). The reproducibility for 1.0 × 10 −13 M target using three electrodes prepared with the same MWCNTs-PolyArg/oligo21 nanoplatform was 5.7%. Figure  We also explored the potential practical application of the biosensor to detect miRNA-21 in real samples such as urine and reconstituted human serum samples. Figure  5C   The calibration plot presented a linear relationship between 10 −14 M and 10 −12 M miRNA-21, with a sensitivity of 1.544 × 10 3 Ω M −1 (r 2 = 0.992), a detection limit (DL) of 3.3 fM, and a quantification limit (QL) of 10 fM (calculated according to 3σ/S and 10σ/S, for DL and QL, respectively, where S is the sensitivity and σ is the standard deviation of the blank signal). The reproducibility for 1.0 × 10 −13 M target using three electrodes prepared with the same MWCNTs-PolyArg/oligo 21 nanoplatform was 5.7%. Figure 5B displays the bars plot for the R ct of the redox marker obtained at GCE/MWCNTs-PolyArg/oligo 21  We also explored the potential practical application of the biosensor to detect miRNA-21 in real samples such as urine and reconstituted human serum samples. Figure 5C displays the bars plot for the R ct obtained at GCE/MWCNTs-PolyArg/oligo 21  A comparison of the analytical performance of our biosensor with the most relevant miRNA-21 electrochemical biosensors reported in the past couple of years is displayed in Table 1. Even when there are some amplified schemes that present lower detection limits than our sensor, it represents an interesting strategy for quantifying miRNA-21 at fM levels in a non-amplified, label-free, fast, and friendly way [33][34][35][36][37][38][39][40][41][42][43][44][45][46][47][48][49].

Conclusions
We report the advantages of MWCNTs-PolyArg nanohybrid for the adsorption and electrooxidation of calf-thymus and salmon sperm double-stranded DNAs. Their most favorable interaction with the aromatic rings of MWCNTs and the positive charge of the polycation that wraps the nanostructures produces a significant decrease in the overvoltages for the oxidation of guanine and adenine residues as well as an important enhancement in the associated currents. This favorable interaction of DNAs with the nanobioplatform allowed us to use MWCNTs-PolyArg as a building block for the construction of a simple, easy-to-prepare, friendly, label-and amplification-free impedimetric genosensor for the highly selective and sensitive quantification of miRNA-21 at femtomolar levels in a very competitive way.