Guanine Quadruplex Electrochemical Aptasensors

Abstract: Guanine-rich nucleic acids are able to self-assemble into G-quadruplex four-stranded secondary structures, which are found at the level of telomeric regions of chromosomes, oncogene promoter sequences and other biologically-relevant regions of the genome. Due to their extraordinary stiffness and biological role, G-quadruples become relevant in areas ranging from structural biology to medicinal chemistry, supra-molecular chemistry, nanotechnology and biosensor technology. In addition to classical methodologies, such as circular dichroism, nuclear magnetic resonance or crystallography, electrochemical methods have been successfully used for the rapid detection of the conformational changes from single-strand to G-quadruplex. This review presents recent advances on the G-quadruplex electrochemical characterization and on the design and applications of G-quadruplex electrochemical biosensors, with special emphasis on the G-quadruplex aptasensors and hemin/G-quadruplex peroxidase-mimicking DNAzyme biosensors.


Introduction
DNA sequences rich in guanine (G) bases are able to self-assemble into four-stranded secondary structures called G-quadruplexes (G4 or GQ), (Scheme 1).The G4s are formed by G-quartet building blocks, which are planar associations of four G bases, held together by eight Hoogsteen hydrogen bonds (Scheme 1B).The G-quartets are stacked on top of each other, stabilized by π-π hydrophobic interactions and by monovalent cations, such as K + and Na + , which are coordinated to the lone pairs of electrons of O6 in each G.
The G4 structures are very polymorphic, being classified according to the number of strands (monomer, dimer or tetramer, Scheme 1C), according to strand polarity (i.e., the relative arrangement of adjacent strands in parallel or antiparallel orientations), glycosidic torsion angle (anti or syn) and the orientation of the connecting loops (lateral, diagonal or both) [1][2][3][4].Different G4 topologies have been observed by nuclear magnetic resonance (NMR) or crystallography, either as native structures or complexed with small molecules [5].
The G4 structures have emerged as a new class of cancer-specific molecular targets for anticancer drugs, since the G4 stabilization by small organic molecules can lead to telomerase inhibition and telomere dysfunction in cancer cells [2,15,16].The G-rich oligonucleotides (ODNs) are also able to self-organize in G4-based two-dimensional networks and long nanowires, relevant for nanotechnology applications [17,18]; therefore, the assembly of G4 nanostructures and devices has been extensively revised in the literature [2,3,5,6,19].
Aptamers are a special class of small synthetic oligonucleotides able to form secondary and tertiary structures, larger than small molecule drugs, but smaller than antibodies, with the advantage of being highly specific in binding to small molecules, proteins, nucleic acids and even cells and tissues [33][34][35][36].The aptamers bind to the targets by a lock and key mode, and the name "aptamer" is from the Latin word aptus, meaning "to fit" [36].Among them, short aptamers that adopt G4 configurations received increased attention, and the electrochemical sensing devices based on G4 nucleic acid aptamers are highly selective and sensitive, fast, accurate, compact, portable and inexpensive.
In this review, recent advances on the G4 structure in nucleic acid electrochemistry and the design and applications of the G4 electrochemical biosensors that use redox labels as amplification strategies, i.e., the G4 electrochemical aptasensors and the hemin/G4 HRP-mimicking DNAzyme electrochemical biosensors, will be presented.the cations that stabilize the G4s are shown as red balls.Adapted from [14] with permission.
The G4 structures have emerged as a new class of cancer-specific molecular targets for anticancer drugs, since the G4 stabilization by small organic molecules can lead to telomerase inhibition and telomere dysfunction in cancer cells [2,15,16].The G-rich oligonucleotides (ODNs) are also able to self-organize in G4-based two-dimensional networks and long nanowires, relevant for nanotechnology applications [17,18]; therefore, the assembly of G4 nanostructures and devices has been extensively revised in the literature [2,3,5,6,19].
Aptamers are a special class of small synthetic oligonucleotides able to form secondary and tertiary structures, larger than small molecule drugs, but smaller than antibodies, with the advantage of being highly specific in binding to small molecules, proteins, nucleic acids and even cells and tissues [33][34][35][36].The aptamers bind to the targets by a lock and key mode, and the name "aptamer" is from the Latin word aptus, meaning "to fit" [36].Among them, short aptamers that adopt G4 configurations received increased attention, and the electrochemical sensing devices based on G4 nucleic acid aptamers are highly selective and sensitive, fast, accurate, compact, portable and inexpensive.
In this review, recent advances on the G4 structure in nucleic acid electrochemistry and the design and applications of the G4 electrochemical biosensors that use redox labels as amplification strategies, i.e., the G4 electrochemical aptasensors and the hemin/G4 HRP-mimicking DNAzyme electrochemical biosensors, will be presented.

G4 Electrochemistry
The first report on the electrochemical oxidation of G4 structures concerned the investigation of two, thrombin-binding aptamer (TBA) sequences, 15-mer d(G 2 T 2 G 2 TGTG 2 T 2 G 2 ) (Scheme 2A) and 19-mer d(G 3 T 2 G 3 TGT 3 T 2 G 3 ) (Scheme 2B), using differential pulse (DP) voltammetry at a glassy carbon electrode (GCE) [32,37].The different adsorption patterns of the TBA sequences observed by AFM onto highly oriented pyrolytic graphite (HOPG) were correlated with their voltammetric behavior in the presence/absence of K + ions.In Na + -containing solutions, the oxidation of both TBA sequences showed one anodic peak corresponding to the oxidation of G residues in the TBA single strands.The G oxidation occurred at the C 8 -H position, in a two-step mechanism involving the total loss of four electrons and four protons.Upon the addition of K + , both sequences folded into G4 structures, causing the decrease of the G oxidation peak current and the occurrence of a new G4 peak at a higher potential, due to the oxidation of G residues in the G4 configuration.In the absence of K + ions, in only Na + ion-containing solutions, G4 formation also occurred, but was much slower.

G4 Electrochemistry
The first report on the electrochemical oxidation of G4 structures concerned the investigation of two, thrombin-binding aptamer (TBA) sequences, 15-mer d(G2T2G2TGTG2T2G2) (Scheme 2A) and 19-mer d(G3T2G3TGT3T2G3) (Scheme 2B), using differential pulse (DP) voltammetry at a glassy carbon electrode (GCE) [32,37].The different adsorption patterns of the TBA sequences observed by AFM onto highly oriented pyrolytic graphite (HOPG) were correlated with their voltammetric behavior in the presence/absence of K + ions.In Na + -containing solutions, the oxidation of both TBA sequences showed one anodic peak corresponding to the oxidation of G residues in the TBA single strands.The G oxidation occurred at the C8-H position, in a two-step mechanism involving the total loss of four electrons and four protons.Upon the addition of K + , both sequences folded into G4 structures, causing the decrease of the G oxidation peak current and the occurrence of a new G4 peak at a higher potential, due to the oxidation of G residues in the G4 configuration.In the absence of K + ions, in only Na + ion-containing solutions, G4 formation also occurred, but was much slower.The decrease of the G oxidation peak current was due to a decrease in the number of free G residues in single-stranded TBA, and the increase of the G4 oxidation peak current was due to an increased number of G4 structures that were more difficult to oxidize.The adsorption of TBA in a G4 conformation, as rod-like-shaped aggregates, was observed by AFM [32,37].
The 10-mer ODNs that contain blocks of 8-10 G residues, d(G)10, d(TG9) and d(TG8T), form parallel tetra-molecular G4s (Scheme 1C, right) and were investigated by AFM and DP voltammetry.The influence of the ODN sequence and concentration, pH (Figure 1), the presence of monovalent cations, Na + vs. K + (Figure 2A,C), and incubation time (Figure 2B,C) was determined [19,32,39,40].The formation of G4 structures and higher-order nanostructures, due to the presence of a long contiguous G region, and the influence of the thymine residues at the 5' and 3' molecular ends in d(TG9) and d(TG8T) were clarified.DP voltammetry allowed the detection of the single strands' association into G4s and G-based nanostructures, in freshly-prepared solutions, at concentrations 10-times lower than usually detected using other techniques currently employed to study the formation of G4s.The decrease of the G oxidation peak current was due to a decrease in the number of free G residues in single-stranded TBA, and the increase of the G4 oxidation peak current was due to an increased number of G4 structures that were more difficult to oxidize.The adsorption of TBA in a G4 conformation, as rod-like-shaped aggregates, was observed by AFM [32,37].
The 10-mer ODNs that contain blocks of 8-10 G residues, d(G) 10 , d(TG 9 ) and d(TG 8 T), form parallel tetra-molecular G4s (Scheme 1C, right) and were investigated by AFM and DP voltammetry.The influence of the ODN sequence and concentration, pH (Figure 1), the presence of monovalent cations, Na + vs. K + (Figure 2A,C), and incubation time (Figure 2B,C) was determined [19,32,39,40].The formation of G4 structures and higher-order nanostructures, due to the presence of a long contiguous G region, and the influence of the thymine residues at the 5 1 and 3 1 molecular ends in d(TG 9 ) and d(TG 8 T) were clarified.DP voltammetry allowed the detection of the single strands' association into G4s and G-based nanostructures, in freshly-prepared solutions, at concentrations 10-times lower than usually detected using other techniques currently employed to study the formation of G4s.       ) 24 h, (E, ‚‚‚) 72 h, (F, ‚ ‚) 5 days and (E, ‚‚‚) 14 days of incubation.Adapted from [39] with permission.Single-stranded ODNs were observed only in Na + ion solutions at short incubation times and were detected in AFM as thin polymeric structures and in DP voltammetry by the occurrence of only the G oxidation peak.The G4 structures were formed very slowly in Na + ions, after a long incubation time, faster in K + ions, after a short incubation time, and were detected by AFM as spherical aggregates and by DP voltammetry by the decrease of the G oxidation peak current and the occurrence/increase of the G4 oxidation peak current, as well as a shift to positive potentials, in a K + ion concentrationand in a time-dependent manner.The presence of K + ions strongly stabilizes and accelerates the G4 formation.For increased d(G) 10 concentrations, long G-nanowires were formed, demonstrating the potential of G-rich DNA sequences as a scaffold for nanotechnological applications [19,32,39,40].
The Tetrahymena telomeric repeat sequence d(TG 4 T) forms parallel-stranded tetra-molecular G4s in the presence of Na + and K + ions and is considered to be a simple model for biologically-relevant G4s.It has also provided high resolution structural data on drug-DNA interactions.The transformation of the d(TG 4 T) from single-stranded into G4 configurations, influenced by the Na + and K + ion concentration, was successfully detected using AFM on HOPG and DP voltammetry at GCE (Figures 3 and 4) [41].The d(TG 4 T) in a G4 conformation self-assembled very quickly in K + ion solutions and slowly in Na + ion solutions.The optimum K + ion concentration for the G4 structure formation of d(TG 4 T) was similar to the intracellular K + ion concentration of healthy cells.In the presence of Na + ions, d(TG 4 T) also formed short nanowires and nanostructured films that were never observed in K + ion-containing solution, suggesting that the rapid formation of stable G4s in the presence of K + is relevant for the good function of cells.
Chemosensors 2016, 4, 13 5 of 20 Single-stranded ODNs were observed only in Na + ion solutions at short incubation times and were detected in AFM as thin polymeric structures and in DP voltammetry by the occurrence of only the G oxidation peak.The G4 structures were formed very slowly in Na + ions, after a long incubation time, faster in K + ions, after a short incubation time, and were detected by AFM as spherical aggregates and by DP voltammetry by the decrease of the G oxidation peak current and the occurrence/increase of the G4 oxidation peak current, as well as a shift to positive potentials, in a K + ion concentration-and in a time-dependent manner.The presence of K + ions strongly stabilizes and accelerates the G4 formation.For increased d(G)10 concentrations, long G-nanowires were formed, demonstrating the potential of G-rich DNA sequences as a scaffold for nanotechnological applications [19,32,39,40].
The Tetrahymena telomeric repeat sequence d(TG4T) forms parallel-stranded tetra-molecular G4s in the presence of Na + and K + ions and is considered to be a simple model for biologically-relevant G4s.It has also provided high resolution structural data on drug-DNA interactions.The transformation of the d(TG4T) from single-stranded into G4 configurations, influenced by the Na + and K + ion concentration, was successfully detected using AFM on HOPG and DP voltammetry at GCE (Figures 3 and 4) [41].The d(TG4T) in a G4 conformation self-assembled very quickly in K + ion solutions and slowly in Na + ion solutions.The optimum K + ion concentration for the G4 structure formation of d(TG4T) was similar to the intracellular K + ion concentration of healthy cells.In the presence of Na + ions, d(TG4T) also formed short nanowires and nanostructured films that were never observed in K + ion-containing solution, suggesting that the rapid formation of stable G4s in the presence of K + is relevant for the good function of cells.Single-stranded ODNs were observed only in Na + ion solutions at short incubation times and were detected in AFM as thin polymeric structures and in DP voltammetry by the occurrence of only the G oxidation peak.The G4 structures were formed very slowly in Na + ions, after a long incubation time, faster in K + ions, after a short incubation time, and were detected by AFM as spherical aggregates and by DP voltammetry by the decrease of the G oxidation peak current and the occurrence/increase of the G4 oxidation peak current, as well as a shift to positive potentials, in a K + ion concentration-and in a time-dependent manner.The presence of K + ions strongly stabilizes and accelerates the G4 formation.For increased d(G)10 concentrations, long G-nanowires were formed, demonstrating the potential of G-rich DNA sequences as a scaffold for nanotechnological applications [19,32,39,40].
The Tetrahymena telomeric repeat sequence d(TG4T) forms parallel-stranded tetra-molecular G4s in the presence of Na + and K + ions and is considered to be a simple model for biologically-relevant G4s.It has also provided high resolution structural data on drug-DNA interactions.The transformation of the d(TG4T) from single-stranded into G4 configurations, influenced by the Na + and K + ion concentration, was successfully detected using AFM on HOPG and DP voltammetry at GCE (Figures 3 and 4) [41].The d(TG4T) in a G4 conformation self-assembled very quickly in K + ion solutions and slowly in Na + ion solutions.The optimum K + ion concentration for the G4 structure formation of d(TG4T) was similar to the intracellular K + ion concentration of healthy cells.In the presence of Na + ions, d(TG4T) also formed short nanowires and nanostructured films that were never observed in K + ion-containing solution, suggesting that the rapid formation of stable G4s in the presence of K + is relevant for the good function of cells.A B ) 200 mM and (‚‚‚) 1 M K + ions.Adapted from [41] with permission.
Synthetic polynucleotides poly(dG) and poly(G) are widely used as models to determine the interaction of drugs with G-rich segments of DNA.AFM and DP voltammetry showed that, at low incubation times, short G4 regions were formed along the poly(G) single-strands, while low adsorption large poly(G) aggregates in a G4 conformation were formed after high incubation times in the presence of either Na + or K + monovalent ions (Figure 5) [42].The DP voltammetry in freshly-prepared poly(G) solutions showed only the G oxidation peak, due to the oxidation of G residues in the poly(G) single strand.Increasing the incubation time, the G oxidation peak current decreased; the peak disappeared; and the G4 oxidation peak in the poly(G) in a G4 conformation appeared, at a higher oxidation potential, depending on the incubation time, presenting a maximum after 10 days of incubation and reaching a steady current after ~17 days of incubation.
Chemosensors 2016, 4, 13 6 of 20 Synthetic polynucleotides poly(dG) and poly(G) are widely used as models to determine the interaction of drugs with G-rich segments of DNA.AFM and DP voltammetry showed that, at low incubation times, short G4 regions were formed along the poly(G) single-strands, while low adsorption large poly(G) aggregates in a G4 conformation were formed after high incubation times in the presence of either Na + or K + monovalent ions (Figure 5) [42].The DP voltammetry in freshly-prepared poly(G) solutions showed only the G oxidation peak, due to the oxidation of G residues in the poly(G) single strand.Increasing the incubation time, the G oxidation peak current decreased; the peak disappeared; and the G4 oxidation peak in the poly(G) in a G4 conformation appeared, at a higher oxidation potential, depending on the incubation time, presenting a maximum after 10 days of incubation and reaching a steady current after ~17 days of incubation.The interaction between the TBA sequences d(G2T2G2TGTG2T2G2) and d(G3T2G3TGT3T2G3) and the serine protease thrombin (Scheme 2) was determined successfully by AFM and voltammetry, taking into account the thrombin interaction with TBA primary and secondary structures, as well as the thrombin folding in the presence of alkaline metals [32,38].In the interaction, the TBA single strands coiled around thrombin, leading to the formation of a robust TBA-thrombin complex that maintained the thrombin symmetry and conformation, which resulted in the thrombin oxidation peaks, within the TBA-thrombin complex, occurring at more positive potentials, than in free thrombin.In the presence of K + , the TBA sequences were folded into a G4 conformation, which The interaction between the TBA sequences d(G 2 T 2 G 2 TGTG 2 T 2 G 2 ) and d(G 3 T 2 G 3 TGT 3 T 2 G 3 ) and the serine protease thrombin (Scheme 2) was determined successfully by AFM and voltammetry, taking into account the thrombin interaction with TBA primary and secondary structures, as well as the thrombin folding in the presence of alkaline metals [32,38].In the interaction, the TBA single strands coiled around thrombin, leading to the formation of a robust TBA-thrombin complex that maintained the thrombin symmetry and conformation, which resulted in the thrombin oxidation peaks, within the TBA-thrombin complex, occurring at more positive potentials, than in free thrombin.In the presence of K + , the TBA sequences were folded into a G4 conformation, which facilitated the interaction with thrombin.The TBA-thrombin complexes adsorbed on the carbon electrode with the TBA in contact with the surface and the thrombin on top, far from the surface; thus, the thrombin molecule was less accessible to oxidation, also leading to the occurrence of the thrombin oxidation peaks at more positive potentials.
A large number of potent G4-binding ligands, which stabilize or promote G4 formation, has been described.Especially at the chromosomes' telomeric regions, the telomeric DNA is able to form G4 structures; therefore, the G4 ligands prevent the G4s from unwinding and opening the telomeric ends to telomerase, thus indirectly targeting the telomerase and inhibiting its catalytic activity.
A number of acridine derivatives have been specifically synthesized with the purpose of increasing binding affinity and selectivity for human telomeric G4 sequences found in chromosomes' telomeric regions, e.g., BRACO-19 [43] and RHPS4.More recently, a new series of triazole-linked acridine ligands, e.g., GL15 and GL7 [44], with enhanced selectivity for human telomeric G4s binding versus duplex DNA binding, have been designed, synthetized and evaluated.
The interactions of the GL15 triazole-acridine conjugate with the short-length Tetrahymena telomeric DNA repeat sequence d(TG 4 T) and with the long chain poly(G) sequence, at the single-molecule level, by AFM and DP voltammetry, were investigated [45].GL15 interacted with both the d(TG 4 T) and poly(G) sequences, in a time-dependent manner, and the influence of Na + vs. K + ions was evaluated.
The G4 formation was detected in AFM, by the adsorption of small d(TG 4 T) and poly(G) spherical aggregates, as well as large G4-based poly(G) assemblies, and the DP voltammetry showed the decrease and disappearance of the GL15 and the G oxidation peak currents and the appearance of the G4 oxidation peak (Figure 6).The GL15 strongly stabilized and accelerated the G4 formation in both Na + and K + ion-containing solutions, although only K + promoted the formation of perfectly-aligned tetra-molecular G4s.The GL15-d(TG 4 T) complex with the G4 configuration was discrete and approximately globular, whereas the GL15-poly(G) complex with the G4 configuration was formed at a number of points along the length of the polynucleotide, analogous to beads on a string.
Chemosensors 2016, 4, 13 7 of 20 facilitated the interaction with thrombin.The TBA-thrombin complexes adsorbed on the carbon electrode with the TBA in contact with the surface and the thrombin on top, far from the surface; thus, the thrombin molecule was less accessible to oxidation, also leading to the occurrence of the thrombin oxidation peaks at more positive potentials.
A large number of potent G4-binding ligands, which stabilize or promote G4 formation, has been described.Especially at the chromosomes' telomeric regions, the telomeric DNA is able to form G4 structures; therefore, the G4 ligands prevent the G4s from unwinding and opening the telomeric ends to telomerase, thus indirectly targeting the telomerase and inhibiting its catalytic activity.
A number of acridine derivatives have been specifically synthesized with the purpose of increasing binding affinity and selectivity for human telomeric G4 sequences found in chromosomes' telomeric regions, e.g., BRACO-19 [43] and RHPS4.More recently, a new series of triazole-linked acridine ligands, e.g., GL15 and GL7 [44], with enhanced selectivity for human telomeric G4s binding versus duplex DNA binding, have been designed, synthetized and evaluated.
The interactions of the GL15 triazole-acridine conjugate with the short-length Tetrahymena telomeric DNA repeat sequence d(TG4T) and with the long chain poly(G) sequence, at the single-molecule level, by AFM and DP voltammetry, were investigated [45].GL15 interacted with both the d(TG4T) and poly(G) sequences, in a time-dependent manner, and the influence of Na + vs. K + ions was evaluated.
The G4 formation was detected in AFM, by the adsorption of small d(TG4T) and poly(G) spherical aggregates, as well as large G4-based poly(G) assemblies, and the DP voltammetry showed the decrease and disappearance of the GL15 and the G oxidation peak currents and the appearance of the G4 oxidation peak (Figure 6).The GL15 strongly stabilized and accelerated the G4 formation in both Na + and K + ion-containing solutions, although only K + promoted the formation of perfectly-aligned tetra-molecular G4s.The GL15-d(TG4T) complex with the G4 configuration was discrete and approximately globular, whereas the GL15-poly(G) complex with the G4 configuration was formed at a number of points along the length of the polynucleotide, analogous to beads on a string.Adapted from [45] with permission.

G4 Electrochemical Biosensors
A DNA-electrochemical biosensor is formed by an electrode (the electrochemical transducer) with a DNA probe immobilized on its surface (the biological recognition element) and is used to detect DNA-binding molecules (the analyte) that interact and induce changes in the DNA structure and electrochemical properties, which are further translated into an electrical signal [25][26][27][29][30][31]46,47].Up to now, the G4-based electrochemical biosensors reported in the literature always used redox labels as amplification strategies.Two important types of G4 electrochemical biosensors, the G4 electrochemical aptasensors and the hemin/G4 DNAzyme electrochemical biosensors, will be revisited.
The first G4 electrochemical aptasensors used TBA sequences and gold electrodes as electrochemical transducers, the aptamers' attachment being achieved using an amine or a thiol functionalization [50][51][52], or the affinity of biotin to avidin, streptavidin or neutravidin [53].Depending on the assay format, two main G4 electrochemical aptasensor categories can be depicted, the sandwich-type aptasensors (also named dual-site binding) and the structure switching-based aptasensors (single-site binding) [54,55].

Sandwich-Type G4 Electrochemical Aptasensor
The aptamer-analyte-aptamer sandwich-type G4 electrochemical aptasensor (Scheme 3A) is composed by two aptamer layers, the first aptamer layer being immobilized on the electrode and used for capturing the analyte and the second aptamer layer being labelled and used for the electrochemical detection.The first aptamer was generally immobilized onto gold via a thiol [56][57][58][59][60] and, more recently, magnetic beads [61,62].The labels of the second aptamer were either redox molecules, nanocomposites [60], nanoparticles [57,63], quantum dots [58,59] or enzymes, with catalytic activity that transformed the substrate into an electroactive product [56,64,65].the analyte is bound to the surface via an aptamer, and a labelled antibody is used for detection.
Employing platinum nanoparticle labels as catalysts for the reduction of H2O2 to a TBA/thrombin complex allowed the amplified electrocatalytic detection of thrombin with a 1 nM detection limit [57].
In another report, gold nanoparticles' functionalization of the second aptamer improved the The first G4 electrochemical aptasensor reported presented an aptamer-analyte-aptamer sandwich-type format, being developed for thrombin detection [64,65].The sensor was built up by two aptamers, the first aptamer immobilized onto the gold electrode for capturing the thrombin onto it and the second one, a glucose dehydrogenase (GDH)-labelled anti-thrombin aptamer.The current increase generated by the electroactive product of the enzyme reaction was observed, and >10 nM thrombin were detected selectively.This approach proved for the first time that aptamers can be successfully employed in sandwich-type sensing devices, instead of and with advantages over antibodies.
Employing platinum nanoparticle labels as catalysts for the reduction of H 2 O 2 to a TBA/thrombin complex allowed the amplified electrocatalytic detection of thrombin with a 1 nM detection limit [57].
In another report, gold nanoparticles' functionalization of the second aptamer improved the thrombin detection sensitivity, showing a 0.02 nM detection limit, with a 0.05-18 nM linear range [63].
The use of cadmium sulfide quantum dot labels of the secondary aptamer allowed thrombin detection with a 0.14 nM detection limit, corresponding to 28 fmol of analyte [58], while in another similar approach, thrombin determination in human serum showed a detection limit as low as 1 pM [59].
In another approach, an aptamer-analyte-aptamer sandwich-type G4 electrochemical aptasensor was based on enzymatic labelling of the second aptamer with glucose dehydrogenase (GDH), measuring the electric current generated by the glucose oxidation catalyzed by GDH and selectively detecting 1 µM of thrombin [64].
Apart from the aptamer-analyte-aptamer sandwich-type G4 electrochemical aptasensor, other design strategies were also employed.The antibody-analyte-aptamer sandwich-type aptasensor consisted of attaching the analyte to the surface via an antibody combined with a labelled aptamer that adopt the G4 configuration for detection (Scheme 3B).This sensor design was used to detect thrombin at a nanogold-chitosan composite-modified GCE, linked with the aptamer via a polyclone antibody.The electrochemical active marker used was methylene blue (MB) directly intercalated in the probing aptamer.The sensor linear response for thrombin was in the range 1-60 nM with a 0.5 nM detection limit [66].
The aptamer-analyte-antibody sandwich-type aptasensor consisted of attaching the analyte to the surface via an aptamer able to form the G4 structure, the detection being performed with a redox-labelled antibody (Scheme 3C) [67].
A sandwiched immunoassay for thrombin used a NH 2 -functionalized-TBA immobilized on gold nanoparticle-doped conducting polymer nanorod electrodes and a ferrocene label bound to an antithrombin antibody [67].The sensor used the electrocatalytic oxidation of ascorbic acid by the ferrocene moiety, presenting a wide dynamic range of 5-2000 ng¨L ´1 and a low detection limit of 5 ng¨L ´1 (0.14 pM) and was tested in a real human serum sample for the detection of spiked concentrations of thrombin.

Structure-Switching G4 Electrochemical Aptasensor
A different category of G4 electrochemical aptasensors is based on the aptamer structural modifications upon binding the analyte from the single-stranded to the G4 configuration, structure-switching G4 electrochemical aptasensor (Scheme 4).This strategy generally involved the direct immobilization of the aptamer on the electrode surface, while the analyte was present in solution.The electrochemical signal amplification was obtained by labelling the aptamer with a redox tag [68,69].

Structure-Switching G4 Electrochemical Aptasensor
A different category of G4 electrochemical aptasensors is based on the aptamer structural modifications upon binding the analyte from the single-stranded to the G4 configuration, structure-switching G4 electrochemical aptasensor (Scheme 4).This strategy generally involved the direct immobilization of the aptamer on the electrode surface, while the analyte was present in solution.The electrochemical signal amplification was obtained by labelling the aptamer with a redox tag [68,69].Scheme 4. Structure-switching G4 electrochemical aptasensors: the aptamer is modifying its conformation after analyte binding: (A) increasing the distance from the redox label to the electrode (signal off); (B) decreasing the distance from the redox label to the electrode (signal on).Adapted from [19] with permission.
The first report on a structure-switching G4 electrochemical aptasensor used a covalently-attached MB-labelled TBA on a gold electrode [68,76].The aptasensor detection type was signal-off (Scheme 4A), i.e., in the absence of the thrombin target, the immobilized TBA remained relatively unfolded, allowing the electron transfer from the MB label to the electrode surface, while after thrombin binding, the formation of TBA in the G4 configuration was induced, which inhibited the electron transfer (Figure 7A).
The sensor was selective enough to detect thrombin directly in blood serum with a 20 nM thrombin detection limit.Similar G4 electrochemical aptasensors for thrombin were developed in parallel, using ferrocene labels [77][78][79].In another design, a beacon aptamer-based biosensor for thrombin showed a linear signal between 0 and 50.8 nM of thrombin, with a 0.999 correlation factor and an 11 nM detection limit [80].Adapted from [19] with permission.
The first report on a structure-switching G4 electrochemical aptasensor used a covalently-attached MB-labelled TBA on a gold electrode [68,76].The aptasensor detection type was signal-off (Scheme 4A), i.e., in the absence of the thrombin target, the immobilized TBA remained relatively unfolded, allowing the electron transfer from the MB label to the electrode surface, while after thrombin binding, the formation of TBA in the G4 configuration was induced, which inhibited the electron transfer (Figure 7A).A bifunctional aptamer-based electrochemical biosensor for the detection of both thrombin and adenosine was developed [81].The TBA was first immobilized on a gold electrode, and then, it was hybridized with an adenosine aptamer and labelled with MB.In the presence of thrombin or adenosine, the aptamer bonded to thrombin or to adenosine instead of MB, and the decrease of the MB peak current was related to the concentration of either thrombin or adenosine.The sensor showed The sensor was selective enough to detect thrombin directly in blood serum with a 20 nM thrombin detection limit.Similar G4 electrochemical aptasensors for thrombin were developed in parallel, using ferrocene labels [77][78][79].In another design, a beacon aptamer-based biosensor for thrombin showed a linear signal between 0 and 50.8 nM of thrombin, with a 0.999 correlation factor and an 11 nM detection limit [80].
A bifunctional aptamer-based electrochemical biosensor for the detection of both thrombin and adenosine was developed [81].The TBA was first immobilized on a gold electrode, and then, it was hybridized with an adenosine aptamer and labelled with MB.In the presence of thrombin or adenosine, the aptamer bonded to thrombin or to adenosine instead of MB, and the decrease of the MB peak current was related to the concentration of either thrombin or adenosine.The sensor showed a 3 nM thrombin and a 10 nM adenosine detection limit.
A signal-on structure-switching G4 electrochemical aptasensor (Scheme 4B) was described [82], with a short MB-tagged oligonucleotide hybridized with both the thrombin-binding portion of the TBA and the DNA sequence linking the aptamer to the electrode.The thrombin binding induced the structural modification of the TBA in a G4 configuration, liberating the 5 1 end of the tagged oligonucleotide to produce a flexible, single-stranded sequence, which allowed the MB tag to react at the electrode surface, increasing the reduction peak current (Figure 7B).Comparing to the signal-off structure-switching G4 electrochemical aptasensors previously described [68,76], the signal-on aptasensor design achieved a current increase of ~300% with a saturated target and a 3 nM detection limit [82].
An impedimetric aptasensor for thrombin detection was described, based on different TBA sequences directly immobilized on the gold electrode and using phosphoramidite synthons for a strong thiolate anchoring of the aptamer and high flexibility [83].In the presence of the [Fe(CN) 6 ] 3´/4ŕ edox probe, the impedimetric aptasensor exhibited high sensitivity, specificity and stability and a 3.1 ng¨mL ´1 (80 pmol/L) thrombin detection limit.
A signal-on G4 electrochemical aptasensor based on co-immobilization of MP-11 and thiol ferrocene-labeled anti-thrombin aptamer, the interaction being detected via a microperoxidase-mediated electron transfer between the ferrocene and the gold electrode surface, was described [84].The system showed a very high sensitivity of 30 fM using electrochemical impedance spectroscopy.Another impedimetric aptasensor for thrombin, based on a layer-by-layer polyamidoamine dendrimer-modified gold electrode [91], showed in the presence of the reversible [Fe(CN) 6 ] 3´/4´r edox couple a linear relationship with the concentrations of thrombin in the range of 1-50 nM and a 0.01 nM detection limit.
More recently, a signal-on electrochemical aptasensor based on target-induced split aptamer fragments' conjunction was described [86].The new design used TBA splinted into two fragments, one attached to the gold electrode and the second one modified with ferrocene, the association of thrombin inducing the association of the two fragments, thus increasing the concentration of ferrocene at the electrode surface.The signal-on electrochemical aptasensor showed a linear range of 0.8-15 nM and a 0.2 nM detection limit.
Another procedure used to improve the sensitivity of the G4 electrochemical aptasensors for the detection of thrombin used an amplification strategy based on the electrochemical active-inactive switch between monomer/dimer forms of carminic acid (CA).The CA was electroactive, while the CA dimers were electrochemically inactive [92].With magnetic enrichment, the sensor showed a 42.4 pM detection limit.
Nanoparticle-based materials, including gold [63,73], platinum [57] and Fe 3 O 4 nanoparticles [93] and quantum dot-coated silica nanospheres [94], were also used as signal amplification strategies for ultrasensitive electrochemical aptasensing.For example, an electrochemical aptasensor based on gold nanoparticles showed a linear range of 0.05-18 nM and a 0.02 nM detection limit [63], while another one based on Fe 3 O 4 -nanoparticles [93] showed a linear response for thrombin in the range of 1.0-75 nM and a 0.1 nM detection limit.
Based on the aptamer conformational change in the presence of K + cations, different electrochemical aptasensors have been developed for selective potassium recognition.The formation of a G4 structure in the presence of K + ions was detected by monitoring the changes on the electron transfer between redox labels and the electrode surface [95][96][97] or by detecting the changes on the interfacial electron transfer resistance [98].The same strategy for specific recognition of other metal ions, such as Tb 3+ , was used [99].
Taking advantage of the ability of thrombin to catalyze the hydrolysis of the peptide (-Ala-Gly-Arg-nitroaniline) to nitroaniline, thrombin was electrochemically detected, by quantifying the nitroaniline reaction product [56].
A different strategy for G4 electrochemical aptasensors used catalysts, such as horseradish peroxidase (HRP) [56,100].In a simple approach, TBA was non-specifically immobilized on the electrode surface, and thrombin was detected using the HRP label, allowing a 3.5 nM detection limit, sufficient for clinical diagnostic of metastatic lung cancer, where the concentration of thrombin level detected was 5.4 nM [56].
An impedimetric biosensor based on a DNA aptamer specific to ochratoxin A (OTA) covalently immobilized onto a mixed Langmuir-Blodgett monolayer composed of polyaniline-stearic acid and deposited on indium tin oxide (ITO)-coated glass plates showed a 0.24 nM detection limit [85].The system was further improved [87], showing a detection limit comparable to that of the HPLC method (0.12 nM), and was validated in food samples.Another design for the OTA detection proposed the use of a long polyethylene glycol spacer chain, which led to the formation of long tunnels at the surface of screen-printed carbon electrodes, with aptamers acting as gates.The aptamer changed configuration after OTA binding, and the peak current decreased [88].
In a different approach, OTA was detected at a G4 electrochemical aptasensor that used a hairpin anti-OTA aptamer and site-specific DNA cleavage of TaqaI restriction endonuclease, as well as a streptavidin-HRP tag, being able to detect as low as 0.4 pg/mL OTA with ultrahigh selectivity [100].

Hemin/G4 DNAzyme Electrochemical Biosensor
Hemin/G4 DNAzyme is an artificial catalytically-active DNA molecule (DNAzyme) that is composed of DNA in the G4 configuration with intercalated hemin molecules.Hemin is an iron-containing porphyrin, whose peroxidase activity increases in the presence of DNA, facilitating the redox reaction between H 2 O 2 and a target molecule (the substrate, e.g., 3,3 1 ,5,5 1 -tetramethylbenzidine, hydroquinone or ferrocene methyl alcohol), which results in the appearance of an oxidized target molecule (the electroactive product), that is electrochemically detected (Scheme 5).
Hemin/G4 DNAzyme electrochemical biosensors represent nowadays one of the most popular building assays of G4-based electrochemical biosensors [101].The most common strategy consists of the modification of the electrode by a hairpin nucleic acid oligonucleotide that contains two sequences, a sequence capable of forming a G4 structure that binds the hemin, used as the amplification strategy, and an aptamer able to specifically bind the analyte, which might form or not a G4 structure.In the presence of the analyte and hemin, the hairpin structures are opened, the hemin/G4 structures are formed on the electrode surface, while the analyte binds to the aptamer.
Since the first report on using hemin/G4 DNAzyme as the electrocatalytic label for amplifying Hemin/G4 DNAzyme electrochemical biosensors represent nowadays one of the most popular building assays of G4-based electrochemical biosensors [101].The most common strategy consists of the modification of the electrode by a hairpin nucleic acid oligonucleotide that contains two sequences, a sequence capable of forming a G4 structure that binds the hemin, used as the amplification strategy, and an aptamer able to specifically bind the analyte, which might form or not a G4 structure.In the presence of the analyte and hemin, the hairpin structures are opened, the hemin/G4 structures are formed on the electrode surface, while the analyte binds to the aptamer.
Later on, more complicated amplification strategies were developed to improve the sensitivity of the hemin/G4 DNAzyme HRP-mimicking activity, such as dual-amplification [121], background noise reduction [122] or autocatalytic target recycling strategies [123].
The glucose oxidase activity was followed by a hemin/G4 DNAzyme electrochemical biosensor, by attaching the enzyme to the electrode surface through the nucleic acid sequence able to form G4s in the presence of hemin [102,124].Then, the glucose oxidase mediated the glucose oxidation to gluconic acid and H 2 O 2 , and the resulting H 2 O 2 was analyzed through its electrocatalyzed reduction by the DNAzyme.
Another electrochemical sensing strategy, based on the G4 DNAzyme for the detection of both adenosines and hydrogen peroxide from cancer cells, was developed [112], which detected the flux of H 2 O 2 released from cells with high sensitivity and showed a 0.1 nM detection limit for ATP.
A hemin/G4 DNAzyme-based impedimetric biosensor was used to detect the environmental metabolite 2-hydroxyfluorene (2-HOFlu) [117], using the hemin/G4 HRP-like activity to catalyze the oxidation of 2-HOFlu by H 2 O 2 , with a 1.2 nM detection limit in water and a 3.6 nM detection limit in spiked lake water samples.The assay was also selective over other fluorene derivatives.
A sandwich-type electrochemical aptamer cytosensor for the detection of HepG2 cells was used [106].On the first approach, the sensor was built up by self-assembling thiolated TLS11a aptamers on the surface of gold electrodes and a G4/hemin/aptamer and HRP-modified gold nanoparticles.The sensor detection range was from 102-107 cells¨mL ´1 and had a 30 cells¨mL ´1 low detection limit.
The system was improved by self-assembling the TLS11a aptamers with gold nanoparticles (AuNPs) on the surface of GCE [107].Hybrid Fe 3 O 4 /MnO 2 /Au@Pd nanoelectrocatalysts, hemin/G4 HRP-mimicking DNAzymes and the natural HRP enzyme efficiently amplified the electrochemical signal through catalyzing the oxidation of hydroquinone (HQ) by H 2 O 2 .This cytosensor provided a better 15 cells¨mL ´1 detection limit, good specificity and stability.
In a different approach, the hemin/G4 DNAzyme electrochemical biosensors took advantage of the hemin/G4 acting both as an NADH oxidase, assisting the oxidation of NADH to NAD + together with the generation of H 2 O 2 in the presence of dissolved O 2 , as well as a hemin/G4 DNAzyme to bioelectrocatalyze the reduction of the produced H 2 O 2 .Initially, this approach was used for the detection of thrombin [125][126][127].More recently, the Pebrine disease-related Nosema bombycis spore wall protein was detected, using the amplification of hemin/G4 DNAzyme functionalized with Pt@Pd nanowires, the electrochemical immunosensor exhibiting a linear range from 0.001-100 ng¨mL ´1 and a 0.24 pg¨mL ´1 detection limit [128].
A DNAzyme that simultaneously served as an NADH oxidase and HRP-mimicking DNAzyme was developed to detect mercury ions (Hg 2+ ) [129], with the dynamic concentration range spanning from 1.0 ng L ´1-10 mg¨L ´1 Hg 2+ and a 0.5 ng¨L ´1 (2.5 pM) detection limit, also demonstrating an excellent selectivity against other interferent metal ions.
A pseudo triple-enzyme cascade electrocatalytic electrochemical aptasensor for the determination of thrombin, using the amplification of an alcohol dehydrogenase (ADH)-Pt-Pd nanowire bionanocomposite and a hemin/G4 structure that simultaneously acted as NADH oxidase and HRP-mimicking DNAzyme, was developed [130].The ADH immobilized on the Pt-Pd nanowires catalyzed the ethanol present in the electrolyte into acetaldehyde, accompanied by NAD + being converted to NADH.Then, the hemin/G4 firstly served as NADH oxidase, converting the produced NADH to NAD + , then the hemin/G4 acting as the HRP-mimicking DNAzyme bioelectrocatalyzed the produced H 2 O 2 .In this way, a concentration linear range from 0.2 pM-20 nM with a low 0.067 pM detection limit for thrombin was obtained.
Another strategy for thrombin detection consisted of using porous platinum nanotubes (PtNTs) labelled with hemin/G4 and GDH [131].Coupling with GDH and hemin/G4 as NADH oxidase and HRP-mimicking DNAzyme, the cascade signal amplification allowed the detection limit of thrombin down to the 0.15 pM level.

Conclusions
The detailed knowledge of G4 formation mechanism, at the surface of electrochemical transducers, is of utmost importance for the design and fabrication of G4-based electrochemical aptasensors, with applications in nanotechnology and biosensor technology.The voltammetric techniques in combination with AFM were successfully employed to study the transformation of single-strand sequences into the G4 configuration or G4-based nanostructures, in freshly-prepared solutions, for concentrations 10-times lower than usually detected by other techniques, such as UV absorbance, circular dichroism or electrospray mass spectroscopy.
The key features of the G4 conformation in nucleic acid electrochemistry and their application in G4 electrochemical biosensors that use redox labels as amplification strategies, i.e., the G4 electrochemical aptasensors and the hemin/G4 HRP-mimicking DNAzyme electrochemical biosensors, were revised.

Scheme 1 .
Scheme 1. (A) Chemical structure of the guanine (G) base; (B) G-quartet and (C) G-quadruplex (G4); the cations that stabilize the G4s are shown as red balls.Adapted from [14] with permission.

Scheme 1 .
Scheme 1. (A) Chemical structure of the guanine (G) base; (B) G-quartet and (C) G-quadruplex (G4); the cations that stabilize the G4s are shown as red balls.Adapted from [14] with permission.
right) 1 M K + ions, 0 h of incubation; (C) AFM images of d(G) 10 in the absence/presence of different K + ion concentrations and different incubation times.Adapted from [40] with permission.

Figure 3 .
Figure 3. AFM images of d(TG4T) in the presence of K + ions, after (A) 0 h, (B) 48 h and (C) 7 days of incubation.Adapted from [41] with permission.

Figure 3 .
Figure 3. AFM images of d(TG4T) in the presence of K + ions, after (A) 0 h, (B) 48 h and (C) 7 days of incubation.Adapted from [41] with permission.

Figure 6 .
Figure 6.GL15-d(TG4T) after different incubation times in the presence of K + ions: (A,B) AFM images and cross-section profiles through the white dotted lines and (C) DP voltammograms baseline corrected.Adapted from [45] with permission.

Figure 6 .
Figure 6.GL15-d(TG 4 T) after different incubation times in the presence of K + ions: (A,B) AFM images and cross-section profiles through the white dotted lines and (C) DP voltammograms baseline corrected.Adapted from[45] with permission.

Chemosensors 2016, 4 , 13 9 of 20 Scheme 3 .
Scheme 3. Sandwich-type G4 electrochemical aptasensors: (A) aptamer-analyte-aptamer sandwich; the first aptamer is used for binding the analyte to the electrode, and the second labelled aptamer is used for detection; (B) antibody-analyte-aptamer sandwich; the analyte is bound to the surface via an antibody, and a labelled aptamer is used for detection; (C) aptamer-analyte-antibody sandwich; the analyte is bound to the surface via an aptamer, and a labelled antibody is used for detection.Adapted from[19] with permission.

Scheme 3 .
Scheme 3. Sandwich-type G4 electrochemical aptasensors: (A) aptamer-analyte-aptamer sandwich; the first aptamer is used for binding the analyte to the electrode, and the second labelled aptamer is used for detection; (B) antibody-analyte-aptamer sandwich; the analyte is bound to the surface via an antibody, and a labelled aptamer is used for detection; (C) aptamer-analyte-antibody sandwich; the analyte is bound to the surface via an aptamer, and a labelled antibody is used for detection.Adapted from[19] with permission.

Scheme 4 .
Scheme 4. Structure-switching G4 electrochemical aptasensors: the aptamer is modifying its conformation after analyte binding: (A) increasing the distance from the redox label to the electrode (signal off); (B) decreasing the distance from the redox label to the electrode (signal on).Adapted from[19] with permission.

Chemosensors 2016, 4 , 13 11 of 20 Figure 7 .
Figure 7. Structure-switching G4 electrochemical aptasensor for thrombin: (A) signal off: thrombin binding reduces the current from the MB redox tag; and (B) signal on: thrombin binding increases the current from the MB redox tag.Reproduced from [68] with permission.

Figure 7 .
Figure 7. Structure-switching G4 electrochemical aptasensor for thrombin: (A) signal off: thrombin binding reduces the current from the MB redox tag; and (B) signal on: thrombin binding increases the current from the MB redox tag.Reproduced from [68] with permission.