After > 50 years of its existence, electrochemistry of NAs is a booming field, currently aimed at developing DNA sensors and sensing assays. A huge amount of knowledge on NA interactions with electrically charged surfaces summarized in this review makes electrochemistry of NAs potentially useful in various fields of biochemical research. DNA and RNA, as well as their mimetics, such as PNA, are electroactive species, producing oxidation and reduction signals of their bases at some electrodes (sections 3.1 and 3.2). Moreover, these NAs can produce capacitive signals related to their adsorption/desorption behavior (section 3.3). Both the faradaic and capacitive signals reflect changes in the DNA structure under conditions close to physiological; highest sensitivity to small structural changes was observed with mercury and solid amalgam electrodes. Using proper EC methods and ionic conditions, either (a) the secondary changes in the DNA structure at the electrode surface can be eliminated to obtain information about the DNA structure in solution, or (b) structural changes can be induced by prolonged contact of dsDNA with electrically charged surface, SSfollowed by their EC detection. Application of negative charges to the surface-attached DNA may result in the DNA denaturation and eventual strand separation at the electrode surface (section 4.2). At positively charged surfaces, no denaturation was observed and stabilization of DNA at these surfaces was reported. However, it is unclear whether DNA assumes a double-helical structure at surfaces or whether the DNA duplex adopts a ladder-like or some other, more or less unwound structure prior to its opening at negative potentials (section 4.4). Label-free methods (based e.g., on the intrinsic electroactivity of NAs) are simple and convenient, but in many cases DNA labeling offers better sensitivity and other advantages. Covalently bound electroactive labels can be easily introduced in NAs (section 5.2). Some labels (such as ferrocene) can be bound to ODNs during their (usually commercial) synthesis in the organic chemistry laboratories. Os(VIII) complexes can be introduced into DNA, RNA, and PNA by addition to the 5,6- double bond in pyrimidine bases, performed just by mixing the reagent with NA at room temperature. Different labels can be also attached to DNA during its enzymatic synthesis. DNA labeling is particularly important for specific end-labeling of target or reporter probe DNAs. In the recent decade, the NA labeling was greatly influenced by application of nanotechnologies (section 5.3). First papers on NA electrochemistry were published >50 years ago, but for about 30 years DNA electrochemistry was a small field involving handful of laboratories, publishing in average ∼10 papers per year. Starting from 1990, an exponential increase in a number of papers occurred, mounting to >700 papers per year during the recent years (Figure 1). This large increase is related to the progress in genomics and particularly in the Human Genome Project, requiring new methods for parallel DNA nucleotide sequencing. EC methods arrived to this field later than optical methods, but their outlook for practical application appear bright, because their performance is now comparable to optical methods; yet EC methods are simpler, less expensive, easily adaptable for miniaturization and well-suited for decentralized analysis and inclusion into LOC. During the first 30 years, the electrochemistry of NAs dealt mainly with basic EC problems, such as electroactivity and adsorption/desorption of NAs, but also with DNA structure in solution and at interfaces (producing early data on DNA premelting and polymorphy of the DNA double helical structure 138 in agreement with trends in the DNA research in that time). In spite of this orientation, many early steps important in the present development of the EC DNA sensors were done. For example, application of solid carbon electrodes, 139 covalent labeling of DNA, 126-128 invention of DNA-modified electrodes, 125 detection of DNA renaturation 100,102,104 and DNA damage 133,840,841 etc., were published before 1990 (Table 2). The development of EC biosensors for DNA hybridization (nucleotide sequencing) started with rather primitive methods using carbon and gold electrodes in combination with redox indicators (binding preferentially to dsDNA). Alternatively, label-free detection based on guanine oxidation signals at carbon electrodes, or later G oxidation with a mediator at ITO electrodes, was used. At gold electrodes, DNA was attached to the surface via its terminal -SH group, forming a SAM with standing-up DNA molecules. At carbon electrodes, unlabeled probe DNA was lying flatly at the electrode surface. These techniques worked relatively well with synthetic ODN targets. They were, however, mostly poorly efficient in the analysis of real DNA samples. To improve the abilities of EC analysis of DNA in biological matrices, about 10 years ago the DST was proposed, in which the DNA hybridization was performed at one surface (usually magnetic beads, optimized for capturing target DNA or RNA from biological materials) and EC detection of the DNA hybridization was done at another surface, that is, at the detection electrode best suited for the given electrode process (section 5.1). DST offered very high sensitivity and specificity in the analysis of real DNA samples, but it required more manipulation than usual SST or an efficient microfluidic system. In the second half of the 1990s, Barton et al. demonstrated unique charge transfer between methylene blue intercalated in the assembly of 15-20 base pair duplex DNAs and a gold electrode to which the duplex was attached via thiol tether (section 5.6). In the following years, this system was improved and employed in the design of various EC assays including DNA hybridization and single-base mismatches. Recently, similar charge transfer has been shown with 100-mer DNA duplex containing covalently bound Nile Blue redox label. Presence of a single base mismatch attenuated the EC signal similarly as in earlier studies of shorter DNA duplexes. In 2003, a new type of a DNA sensor (called E-DNA sensor) was proposed in Heeger's laboratory, based on a change in the structure of ferrocene-labeled DNA hairpin probe into a linear duplex, resulting from the DNA hybridization, resembling thus the molecular beacons based on optical detection (section 5.5). In the hairpin probe, the ferrocene label was located close to the electrode surface and produced an EC signal. Upon the interaction with complementary target DNA, the hairpin changed into a duplex and the label was moved away from the surface, diminishing the EC signal. Later, this signal-off technique was improved and transformed into a more versatile, signal-on technology. Most of the above techniques are strongly dependent on the nature of the electrode used for the DNA sensing. In the ECPs, this dependence is less strict and the performance of the ECP sensor is more dependent on the nature and way of ECP polymerization. ECP may play a passive role, serving just for DNA immobilization, but it can also directly influence the transduction process, manifested by a change of the ECP conductivity, redox behavior, etc. (section 5.4). In such cases, using electrically neutral PNA as a probe is very convenient, because PNA binding to negatively charged target DNA results in a large, easily detectable change in the electrical properties of the DNA-PNA duplex. Papers on development of the DNA sensors dealing with synthetic ODN targets displayed reasonable performance. Real EC analysis does not, however, work with such ODNs and analysis of genomic DNA sequences mostly requires amplification of tDNA by PCR. Compared to synthetic ODN targets, PCR amplicons are usually longer and may contain some additional substances, such as nucleotides and proteins. A number of EC methods have been developed, suitable for the analysis of PCR-amplified DNA and RNA (section 6.1). Moreover, attempts have been made to use EC analysis in endpoint detection of PCR amplicons (section 6.1.1) and to replace optical detection in real-time PCR by EC detection (section 6.1.3). Analysis of NAs without PCR amplification is much more difficult than the detection of PCR amplicons and represents a challenge. In recent years, significant progress has been done, including analysis of unamplified uropathogen rRNA, as well as messenger and microRNAs (section 6.3), frequently based on combination of EC and biochemical approaches. Analysis of unamplified real NA samples is rather difficult because it has to be done in complex biological matrices, such as cell culture, blood, saliva, or urine and requires very high sensitivities and signal-to-noise ratios. S/N has been recently greatly increased by improved shielding of gold electrode surfaces by binary and ternary SAMs. Also DST has shown good properties in the NA analysis in complex biological matrices. Combination of efficient shielding of the surface (at which tDNA is captured) with sandwich assay using TMB as a substrate for HRP/H 2O 2 oxidation appear now very useful in analysis of PCR-unamplified, biologically relevant NA samples (section 6). Cytosine methylation plays important roles in various diseases, including cancer. Simple DNA hybridization techniques cannot be used to detect methylated cytosine, because both base residues exhibit the same base pairing behavior.
