June 23, 2007 12:9 WSPC/204-BRL 00047

Biophysical Reviews and Letters, Vol. 2, No. 2 (2007) 167–189c© World Scientific Publishing Company

RECENT TRENDS IN DNA BIOSENSING TECHNOLOGIES

ELICIA L. S. WONG

Physical and Theoretical Chemistry LaboratoryUniversity of Oxford, South Parks Road

Oxford, OX1 3QZ, United Kingdomelicia.wong@chem.ox.ac.uk

Received 21 July 2006Revised 2 November 2006

Recent development and challenges in DNA biosensing technology for the detection ofDNA hybridization are reviewed with respect to their abilities to achieve lower detec-tion limit and higher selectivity. Researchers exploit a range of different chemistries forthe development of DNA hybridization biosensors, however all the designs take advan-

tage of heterogenous hybridization between the surface-bound DNA (the probe) andthe DNA sample (target) in the solution. The detection protocols include using opti-cal, microgravimetry, and electrochemical-based device to transduce DNA hybridizationby observing changes in light, mass/frequency, and current/charge, respectively, uponexposure to the sample. The pros and cons of these biosensor designs are discussed withillustrative examples.

Keywords: DNA biosensor; DNA hybridization; electrochemical; microgravimetry; label-free; intercalator; enzyme; metal-nanoparticle.

1. Introduction

Over the past decade, DNA sensing has become increasingly important owing to theadvances in medical technologies which lead to the discovery of more genetic dis-eases. As a result, DNA biosensor technologies that focus on the direct detection ofnucleic acids are currently an area of tremendous interest as they can be employedto detect the presence of genes or mutant genes associated with genetic/infectiousdiseases rapidly and efficiently.1 The DNA sensing technology relies on heteroge-neous hybridization between a surface-bound DNA with known genomic sequence,i.e., the probe, and the free target DNA with complementary genomic sequencein the solution. This new technology has the potential to replace the conventionaltechniques such as southern blotting2 which is usually slow and requires hours todays to produce a reliable outcome. However, arrays of DNA can offer high accuracyof multiplex analysis of different genomic sequences. For example, the GeneChip

developed by Affymetrix3 is able to provide DNA sequence information with anassay time of 16 h via hybridization of labeled DNA in solution to DNA molecules

167

June 23, 2007 12:9 WSPC/204-BRL 00047

168 E. L. S. Wong

immobilized at a specific location on the chip. Such a DNA microarray is an excel-lent research tool but is less compatible with rapid monitoring of specific DNAsequence especially mutated sequences with single nucleotide polymorphism (SNP),outside the research laboratory.

For DNA hybridization biosensors to be commercially viable for the analysis ofbiological samples, there are several requirements that need to be satisfied. Firstof all, the biosensor requires an easy-to-construct sensing interface which is able togive a reproducible assay signal. Second of all, the transduction methods need to be(i) highly specific toward the detection of a specific genomic sequence to an extentthat it allows the detection of SNP; (ii) highly sensitive with high signal-to-noiseratio to minimize the likelihood of a false-positive outcome; (iii) economic whichrequires no or little expensive labeling materials on either probe or target DNA; and(iv) user friendly such that no skilled personnel are required. Lastly, the prospectof DNA sensor array development needs to be fulfilled so that detection of multiplegenomic sequences can be carried out simultaneously.

Currently, the common transducing elements for DNA biosensors include theuse of optical,4–9 mass-sensitive10 and electrochemical11–15 devices, which uti-lizes light, frequency, and current/charge, respectively, to detect DNA hybridiza-tion. Different DNA immobilization methods are exploited in attaching theprobe DNA onto the surface of the transducer, for example by adsorption,16–20

copolymerization,21,22 complexation4 and covalent attachment13–15,23,24 depend-ing on the chemical characteristic of the solid support, which can be a metal, glass,or carbon surface.

This article reviews the recent trends and provides an insight to the ongoingresearch on inventive designs of different types of DNA hybridization biosensors.Illustrative examples are included to compare and contrast these biosensor designs.

2. Optical DNA Hybridization Biosensors

Different types of optical DNA hybridization biosensors have been developed. Theseoptical techniques involve the use of fluorescent or surface plasmon resonance (SPR)spectroscopy depending on whether a fluorescent label is employed.

A typical fluorescent DNA biosensor relies on the emission signal from afluorescent-label which is bound to either the DNA duplex or target DNA to trans-duce hybridization via the use of a fluorometer. Normally, fiber optics are used asthe medium to transduce this signal as they allow light transmission by series ofinternal reflections. In one example, Krull and co-workers5 covalently immobilizedss-DNA onto the end of an optical fiber and subsequently hybridized with comple-mentary target DNA in the solution. Hybridization on the probe DNA modifiedoptical fibers was detected using ethidium bromide as the fluorescent label and itcan only fluoresce upon intercalation into a DNA duplex which is not possible withss-DNA. Since this sensing approach essentially relies on the ability to accumulateethidium bromide through intercalation into DNA duplexes, the aspects of discrim-ination against base pair mismatches is limited as the presence of mismatches will

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 169

not greatly affect its ability to form a DNA duplex. However, the prospect of usingthis technology within the design of a DNA sensor array can be met as shown byWalt and co-workers25 where they demonstrated simultaneous detection of mul-tiple DNA genomic sequences using fiber-optic DNA array. In this example, thehybridization of a fluorescently labeled complementary target was monitored byobserving an increase in fluorescence that accompanied binding.25

Fluorescent-labeling of the probe DNA strands eliminates the need to modifythe target strand which allows direct fluorescent detection of DNA hybridizationthat can offer higher sensitivity and specificity. This is demonstrated through aninnovative approach with the design of molecular beacons as shown in Fig. 1.9 Thebiotinylated probe ss-DNA molecular beacon was immobilized onto a solid silicasurface through biotin–avidin binding. The molecular beacon consists of a stem-and-loop structure, with the loop being a sequence of probing ss-DNA and thestem made up of short complementary sequences of DNA at each end of the probecausing the fluorophore (tetramethylrhodamine) to be located next to the quencher(dimethylaminoazobenzen aminoexal-3-acryinido). When the probe sequence in theloop hybridizes with the complementary target in the solution, the formation of rigidDNA duplex separates the quencher from the fluorophore, restoring the fluorescentstatus of the fluorophore.

Fig. 1. Schematic operation of molecular beacons, consisting of biotinylated probe ss-DNA instem-and-loop structures. The fluorophore becomes fluorescent upon hybridization as the quencheris separated from the fluorophore. Q = quencher, F = fluorophore.

June 23, 2007 12:9 WSPC/204-BRL 00047

170 E. L. S. Wong

Another optical transduction method, SPR offers label-free in situ detection ofDNA hybridization. SPR is an evanescent wave optical technique which is sensitiveto changes in the surface optical properties, e.g., a shift in resonance angle of thereflectivity of the light at the transducer surface upon hybridization of the probe andthe target in the solution. It reports changes in refractive index of a thin recogni-tion layer on a metal substrate that occurs upon adsorption of the ss-DNA and alsohybridization with the target DNA. Watt and co-workers26 further combined a reso-nant mirror to the SPR to improve the sensitivity of this type of optical sensor evenfurther. A two-color SPR was also utilized to study the kinetics of hybridization asit allowed unique determination of both the thickness and dielectric constant of thesensing surface.27

Mirkin and co-workers developed a straightforward optical readout approachwhich involves labeling probe DNA with gold nanoparticles that change color fromred to purple upon hybridization with the target DNA. This method is desirable asit allows rapid detection and it provides a colorimetric response with good selectivitywhich requires little or no instrumentation.28,29

3. Microgravimetry DNA Hybridization Biosensors

As with using SPR for DNA sensing, the microgravimetry DNA biosensor isalso able to offer label-free in situ detection of DNA hybridization throughacoustic waves, surface acoustic waves or Love waves. Acoustic wave detectionusing the quartz crystal microbalance (QCM) has been demonstrated by manyresearchers.30–34 The QCM is known as an extremely sensitive mass-measuringdevice as its resonance frequency decreases with an increase in mass on the QCM.35

A typical configuration of a QCM DNA biosensor consists of probe DNA immobi-lized on the metal plates of a quartz crystal (Fig. 2).

The DNA hybridization is monitored via a decrease in the in situ oscillatingfrequency as a result of the increase in mass associated with the hybridization event.A high specificity DNA QCM biosensor that enables the detection of single-baseDNA mismatches is obtained when the probing DNA is replaced with a peptidenucleic acid (PNA) probe.36 For example, a DNA QCM sensor exhibited a 26–31% decrease in resonant frequency when mismatch target DNA was injected,32

whereas, Wang and co-workers36 showed that no decrease in the resonant frequencywas observed upon addition of mismatch sequence to the PNA QCM biosensor.PNA is an analogue of DNA in which the backbone is a pseudopeptide ratherthan a sugar. It mimics the behavior of DNA and binds to complementary DNAsequence. The neutral backbone of PNA results in stronger binding and greaterspecificity towards a complementary DNA sequence than the usual DNA–DNAhybridization. The sensitivity can also be improved by incorporating anti-ds-DNAantibody, which binds specifically to ds-DNA in the biosensor design. Amplificationof the system and confirmation of the primary formation of the DNA duplexesare achieved through further binding of a secondary antibody (anti-mouse FC-antibody) that binds specifically to anti-ds-DNA antibody.37

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 171

Fig. 2. Immobilization of thiolated probe ss-DNA on gold quartz crystal and subsequenthybridization with target complementary ss-DNA. The thiolated probe is immobilized onto thegold crystal via gold/sulfur self-assembly chemistry.

Besides its ability to detect hybridization, QCM has also been employed toobtain quantitative measurements, such as the surface density of the probe DNA38

and degree of hybridization (i.e., hybridization efficiency)33,38 as real-time frequencyresponses can be obtained. These biosensors have also been used in the studies ofother DNA interactions, such as in situ monitoring of protein-DNA39 and anticancerdrug-DNA bindings.40

4. Electrochemical DNA Hybridization Biosensors

Frequently, the optical and microgravimetry approaches to DNA transduction relyon changes in the amount of DNA at the transducer interface to infer DNAhybridization rather than transducing the hybridization event directly as informa-tion regarding how well the binding between the probe and complementary targetcannot be obtained through the output signals. Regardless of whether the targetDNA is labeled, an increase in the resultant analytical signal indicates an increasein the amount of DNA, which may arise either from hybridization with the probeDNA or non-specific adsorption of target DNA onto the transducer surface. In thecase of optical DNA biosensors, a false-positive diagnosis of hybridization might

June 23, 2007 12:9 WSPC/204-BRL 00047

172 E. L. S. Wong

occur as a result of non-specific adsorption of a fluorescent label onto the trans-ducer. In fact, the incorporation of a molecular beacon9 into the design of an opticalDNA biosensor is the only known optical transduction method that does give infor-mation directly regarding duplex formation as the fluorophore can only fluoresceonce the stem is broken apart. However, one potential limitation of the molecularbeacon approach is that the thermodynamic equilibrium needs to favor the duplexformation in the loop sequence such that the duplex in the stem can be brokenapart to restore the fluorescent status of the fluorophore.

Electrochemical system has distinctive advantages over the optical and micro-gravimetry sensing systems, as it offers a simple, rapid, low cost point-of-care detec-tion for selected target DNA and is suitable for fabrication of miniaturized devices.An impressive number of inventive designs for electrochemical DNA sensing haveemerged. Recently published review articles by Kerman et al.,41 Drummond et al.,42

Wang43 and Gooding44 summarized the state-of-the-art and recent trends in thisbiosensor technology. The most common strategy for electrochemical detection ofhybridization is through the use of a redox-active label where there is a change inaffinity of the redox molecule toward the probe ss-DNA modified interface comparedwith after exposure to the sample with target DNA. The labels range from redox-active DNA specific molecules, e.g., DNA groove binders45 and intercalators,14,15,46

biological molecules such as enzymes47−49 or metal-nanoparticles50−52 such as cad-mium sulfide. An alternative to using redox-active labeled systems is the label-freeapproach which relies on either the intrinsic redox-active properties (e.g., direct oxi-dation) of DNA bases (guanine or adenine)53–56 or a change in electrical propertieson the transducer surface.57

4.1. Label-free electrochemical DNA biosensors

Numerous label-free electrochemical detections of DNA hybridization have beeninvestigated in the past few years through direct oxidation of DNA. Palecek pio-neered the study of electroactivity of DNA in 1960.58 Among the constructs ofDNA, only the DNA bases were found to be electroactive and able to undergo reduc-tion and/or oxidation on the surface of the transducer. The DNA bases undergoreduction at highly negative potentials, which are attainable only using mercuryelectrodes while oxidation occurs at highly positive potentials attainable only usingcarbon electrodes. Typical transduction methods of label-free electrochemical DNAbiosensors rely on monitoring the oxidation current of DNA purine bases, guanineor adenine, as they are the most electroactive DNA bases and can be easily absorbedand undergo oxidation on carbon surfaces. The oxidation currents of guanine andadenine are reported at to be around 1.0 V and 1.3 V versus Ag|AgCl, respectively,in 0.5M acetate buffer solution at pH 4.8.59

A typical electrochemical label-free DNA biosensor consists of probe ss-DNAabsorbed on carbon surfaces and the magnitude of the guanine oxidation currentis recorded prior to hybridization. For example, Wang et al.60 immobilized probe

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 173

ss-DNA onto screen-printed carbon electrodes and used the decrease in guanineresponse of the immobilized probe upon exposure to a guanine-free complementarytarget to monitor hybridization events. However, measurement of the decreased gua-nine oxidation current for the detection of hybridization is very limited as it cannot beused for detecting targets containing guanine bases. Such a limitation is overcome byusing an inosine-modified (guanine-free) probe onto the carbon surface. Like guanine,the inosine moiety binds preferentially with the cytosine base61 but its electroactiv-ity is about three orders of magnitude lower than guanine.62 This inosine-modifiedapproach was utilized by Mascini and co-workers53 in conjunction with polymerasechain reaction (PCR) to detect DNA sequences related to the apolipoprotein E byimmobilizing inosine-modified probes onto screen-printed carbon electrodes, wherethe guanine oxidation current is detected using Osteryoung square wave voltammetry(OSWV). Using the same inosine approach on carbon paste electrodes, Wang et al.56

eliminated the need for PCR by utilizing chronopotentiometry instead of OSWV asthe electrochemical technique to detect hybridization. Chronopotentiometry is effec-tive in discriminating against background contribution at relatively high potentials,therefore gives a more defined guanine oxidation current and the sensitivity is there-fore enhanced by this approach.60,63 Another strategy to improve the signal-to-noiseratio was demonstrated by Wang et al.,64 where a two-step approach was used forcapturing target DNA using probe DNA immobilized onto magnetic beads. The mag-netic beads were separated from the analyte solution after hybridization and wereexposed to acidic solution to depurinate the hybridized DNA on the beads. The freeguanine and adenine DNA bases were collected, analyzed, and amplified using anodicstripping voltammetry in the presence of copper ions64 since Shiraishi and Takahashihave shown previously that the anodic response of free adenine and guanine is greatlyenhanced in copper solution.65

A novel approach to further amplify the guanine oxidation current is through theuse of multi-walled carbon nanotubes (MWNTs).66 A significantly enhanced gua-nine oxidation current is observed when the direct electrochemistry of guanine isperformed at a MWNT modified glassy carbon electrode compared to an unmodifiedglassy carbon electrode using cyclic voltammetry (CV).67 Wang et al.66 reportedan 11-fold increase in the guanine oxidation current using an end-functionalizedMWNT-DNA modified glassy carbon electrode for the detection of DNA sequencesrelated to the breast cancer BRCA1 gene compared to a MWNT-free glassy car-bon electrode. To further enhance the guanine oxidation current, Kerman et al.68

incorporated a combination of sidewall- and end-functionalized MWNT into theirdesign of a label-free DNA biosensor as shown in Fig. 3. The promising conductivityof MWNTs, and the creation of a larger surface area for DNA immobilization bysidewall- and end-functionalized MWNTs further lowered the detection limit downto levels which are compatible with the demand of genetic tests.68

Another amplification approach was reported by Pedano et al.69 through theuse of carbon paste nanotube electrodes for adsorption of probe DNA to detecthybridization via electrochemical oxidation of guanine. They reported a 61-fold

June 23, 2007 12:9 WSPC/204-BRL 00047

174 E. L. S. Wong

Fig. 3. A novel DNA-directed attachment of MWNT onto the carbon surface to amplify theguanine oxidation current for the label-free DNA biosensor which relies on the guanine oxidationto detect hybridization. The sidewall and end of MWNT is functionalized and modified withadenine probes, which are allowed to bind to the thymine probe on the carbon surface. After theDNA-directed attachment of MWNT, an inosine substituted thymine capture probe is allowedto bind to adenine modified MWNT surface, which leaves the inosine substituted part free tohybridize with the target DNA containing guanine. Upon hybridization with target DNA, theappearance of the guanine oxidation current is monitored to detect hybridization.

increase in the guanine oxidation current using a carbon paste nanotube modi-fied electrode compared to a bare carbon paste electrode.69 Such an amplificationclearly demonstrated the advantages of using carbon paste nanotubes as biosensingmaterials.

Therefore, the strength in probing hybridization through direct electrochemistryof guanine is the ease of preparation of the sensing interface as no labeling step isrequired and it is amenable to a range of different transducer surfaces. However,the guanine signals are difficult to discriminate against the high background signalssince the oxidation of guanine occurs at high electrochemical potentials. Besides,this approach will destroy the analyte sample upon assay as the oxidation of guanineis irreversible. As a result, the approach is also not practical for the development ofa DNA sensor array as the guanine signals are difficult to be differentiated amongprobe strands of different genomic sequences.

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 175

Besides using direct oxidation of the guanine bases to detect DNA hybridiza-tion, there are other label-free electrochemical biosensors that rely on the electricalcharge flow through DNA for detection of hybridization. Fink and Schonenbergerreported the first electrical measurement on DNA which is at least 600 nm long.70

The resistivity values obtained from these measurements are comparable to thoseof conducting polymers. The change in electrical property from ss-DNA to ds-DNAforms the basis for this alternative type of label-free DNA transduction. Thiswas demonstrated by Kraatz and co-workers71 where the difference in resistancebetween a metallated DNA duplex (M-DNA) and a normal duplex DNA (B-DNA) was utilized to detect hybridization and single-base DNA mismatch usingelectrochemical impedance spectroscopy. M-DNA is a metallated form of DNAwhich forms a complex with Zn2+ at pH 8.5.72 Its formation causes significantchanges in the electronic properties of the DNA that are readily detected by elec-trochemical impedance spectroscopy. The difference in charge transfer resistancesbetween B- and M-DNA decreased from 190 Ω cm2 for a perfectly matched duplexto 95, 30 and 85 Ω cm2 for single-base mismatch duplex at the top (distal), mid-dle, and bottom (proximal) positions of the monolayer with respect to the goldsurface.71

Changes in the structural conformation of DNA from ss-DNA to ds-DNA wasalso used to transduce hybridization directly without any label, as ss-DNA is afloppy strand while ds-DNA forms a rigid rod-like strand. For example, Korri-Youssoufi et al.73 developed direct electrical detection of DNA hybridization bymonitoring changes in the conductivity of conducting polymer molecular interfacescaused by the rigidity of the DNA duplex upon hybridization. They attached probess-DNA onto a polypyrrole modified transducer surface and a decrease in currentof the electroactive polypyrrole at −0.2 V versus SCE in cyclic voltammogramand a shift to more positive potential of the oxidation wave was used to transducehybridization. Such a change in electrical properties of polypyrrole was attributedto conformation changes of probe ss-DNA into a bulky and stiff DNA duplex whichlimited the degree of freedom of the polypyrrole to transform into a quinoid struc-ture. Changes in double layer capacitance of probe ss-DNA interface prior to andafter hybridization observed in cyclic voltammogram was also used to detect DNAhybridization.57 A greater capacitance was observed for ds-DNA interface as therod-like structure of the duplex opened the access of ions onto the transducer surfacewhile the ss-DNA hindered the access.57

4.2. Redox reporter labeled electrochemical DNA biosensors

For a typical assay using a redox-active labeled system, the immobilized probess-DNA is exposed to the label before exposure to the target DNA. In most caseshybridization is transduced by a change in the magnitude of the label’s electro-chemical current after hybridization. For example, Millan and Mikkelsen13 attachedprobe ss-DNA to glassy carbon electrodes and exposed this in the cationic reporter

June 23, 2007 12:9 WSPC/204-BRL 00047

176 E. L. S. Wong

label, tris(1,10-phenanthroline) cobalt (III) perchlorate Co(phen)3+3 . It showedenhanced electrochemical current in cyclic voltammogram upon hybridization andthis was used to transduce hybridization. The enhanced current is a consequence ofCo(phen)3+3 being a minor groove binder and hence has a greater affinity toward therigid ds-DNA modified surface (compared to ss-DNA surfaces). The work of Ozsozand co-workers74,75 exploited a decrease, rather than an increase, in the affinity ofthe redox reporter to the ds-DNA interface upon hybridization with probe DNA.They used methylene blue (MB) as the label, which has been shown to bind to gua-nine bases76 for DNA transduction. Hybridization reduces the accessibility of theguanine to MB as a result of stronger binding affinity to the complementary targetDNA bases and hence there is a concomitant decrease in MB current observed inOSW voltammograms. These approaches still infer rather than transduce hybridiza-tion as they still rely on a change amount of DNA at the recognition interfaces, i.e.,MB can bind to the unhybridized target strand which adsorbed onto the surface ofthe transducer.

Fan et al.77 also used the decrease in redox reporter current to transduce DNAhybridization. They employed the stem-and-loop molecular beacon biosensor designas illustrated in the design of the optical DNA biosensor as shown in Fig. 1, butthe fluorophore was replaced by an electroactive ferrocene label. The stem-and-loop DNA structure was self-assembled onto a gold electrode by means of the facilegold-thiol chemistry. In the absence of target DNA, the stem-and-loop structureheld the ferrocene label into close proximity with the electrode surface ensuringrapid electron transfer and hence efficient electrochemistry of the ferrocene labelwas observed in cyclic voltammogram (Fig. 4). Hybridization induced a large confor-mational change in the surface-confined DNA and significantly altered the electrontransfer tunneling distance between the electrode and the ferrocene label, causinga decrease in observed ferrocene current.77 This design will give no response withnon-specific adsorption of either DNA or redox label as it relies on the formation ofa perfectly matched duplex which, in turn, transduces hybridization. However, it isnot well-suited for detection of mismatches as mismatch sequences are still capableof forming duplexes with probe strands.

Clinical Microsensors,78,79 a biochip company, has also utilized the ferrocenelabel for transduction of DNA through a three-component “sandwich” hybridiza-tion assay. This electrochemical sandwich assay consists of an immobilized captureprobe, signaling probe and target DNA (Fig. 5). Capture of the signaling probe,which is labeled with a ferrocene, at the DNA biosensor surface allows an electro-chemical current to be recorded using molecular wires. However, the signaling probecan only be captured once hybridization of the target DNA has occurred with theimmobilized capture probe. As the target DNA is longer than the capture probe,the signaling probe hybridized to the “sticky” unhybridized end of the target DNA.This dual hybridization approach eliminates the need to modify the target DNAsample and provides enhanced selectivity from the existence of two hybridizationevents.

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 177

Fig. 4. The stem-and-loop DNA biosensor design with a ferreocene-label.

4.3. Enzyme labeled electrochemical DNA biosensors

In an effort to improve the sensitivity of the hybridization events, a transducersurface can also be modified with a polymer layer which is electrically conducting.Caruana and Heller48 have reported the use of enzyme-amplified DNA biosensors byimmobilizing probe ss-DNA onto an electron-conducting polymer redox surface ona microelectrode and labeling the target DNA with an enzyme, soybean peroxidase(SBP). When the redox polymer and SBP were brought in to close proximity byhybridization, the modified interface switched from being a noncatalyst to a catalystfor electroreduction of hydrogen peroxide at −0.06V versus Ag|AgCl as shown inFig. 6. This electroreduction current of hydrogen peroxide was monitored to detecthybridization.

A similar strategy was also employed in a sandwich-type enzyme-amplifiedfashion, similar to the ferrocene-labeled electrochemical sensor design shown inFig. 5, to further improve the sensitivity of the enzyme labeled electrochemicalDNA biosensors.49,80 The redox polymer and probe ss-DNA were electrodepositedonto a screen-printed electrode, where the probe DNA hybridized with two targetsequences, the analyte and horseradish peroxide (HRP) labeled reporter sequences,as portrayed in Fig. 7. After hybridization, the HRP labels are in electrical contactwith the redox polymer and are able to undergo electrocatalytic reduction of H2O2

to H2O, and the electroreduction current at 0.10V versus Ag|AgCl was measuredto transduce hybridization.80

Multiple amplifications using a sandwich-type of enzyme labeled biosensor werealso achieved by using a liposome as illustrated in Fig. 8.81,82 The employment ofliposome eliminates the need to use a redox polymer to amplify the DNA transduc-tion signal. This sensing method does not rely on the electroreduction current of

June 23, 2007 12:9 WSPC/204-BRL 00047

178 E. L. S. Wong

Fig. 5. Electrochemical “sandwich” assay consisting of a signaling probe, capture probe andtarget DNA.

H2O2 to detect DNA hybridization; instead the appearance of an insoluble prod-uct as a result of the biocatalytic precipitation is used to detect hybridizationvia AC impedance spectroscopy.82 In a variation on this approach, Patolsky etal. further extended the application for the detection of single-base DNA muta-tion using avidin-alkaline phosphatase as a conjugate where its association withthe detector interface catalysed the oxidative hydrolysis of 5-bromo-4-chloro-3-indoylphosphate to the insoluble indigo derivative upon exposure to single-baseDNA mutated sequence.83,84

4.4. Metal-nanoparticle labeled electrochemical DNA biosensors

Colloidal metal-nanoparticles have also been used to transduce hybridization. Thebasic protocols for the detection of hybridization using metal-nanoparticles are illus-trated in Fig. 9. There are currently three approaches toward DNA detection usingmetal-nanoparticles: (i) the electrochemistry of the metal-nanoparticle is observedon the DNA-modified electrode52; (ii) on a bare transducer after dissolving the

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 179

Fig. 6. Schematic diagram and detection scheme of the electrochemical DNA biosensor labeledwith enzyme. The probe DNAs were covalently bound to the electron-conducting redox polymeron the transducer. The enzyme labels (SBP) were covalently bound to the target DNA. Uponhybridization, electrical contact was established between the SBP heme centers and the transducervia the redox polymer. This contact enabled the electrocatalytic reduction of added substrates,H2O2, to the products, H2O, through the cycle. The current of H2O2 electroreduction was therebyused to transduce DNA hybridization.

metal-nanoparticles with acid treatment,85 and (iii) the metal-nanoparticle is sil-ver coated and the electrochemistry of silver is detected with or without silverdissolution.50,86 A recent example of this approach was demonstrated by Ozsozet al.,52 who attached target DNA onto a pencil graphite electrode (PGE). WhenPGE modified target DNA hybridized with complementary probe DNA, which was

June 23, 2007 12:9 WSPC/204-BRL 00047

180 E. L. S. Wong

Fig. 7. Sandwich-type enzyme amplified assay of DNA hybridization.

conjugated to a gold-nanoparticle, the hybridization was detected electrochemicallywith the appearance of a gold-oxidation current using pulse-voltammetry. The cur-rent was greatly enhanced because of the availability of many oxidizable gold atomsand the relatively large surface area in each nanoparticle label.

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 181

Fig. 8. Amplified detection of target DNA by biotion-labeled-HRP-functionalized liposomes andbiocatalyzed precipitation of an insoluble product. A = Avidin, B = Biotin, P = Precipitate.

As potentiometric stripping voltammetry is known to be a powerful techniquefor trace metal measurements, the dissolution of metal-nanoparticles from thehybridized DNA strands for stripping analysis offers higher sensitivity toward DNAdetection. This was demonstrated by Wang et al.85 where they successfully detectedthe breast cancer DNA target using chronopotentiometric stripping analysis of thedissolved gold colloid from the hybridized DNA strand. Further amplification andlowering of the detection limits were achieved either by using a silver-nanoparticle86

or by coating silver onto the gold-nanoparticle after oxidative dissolution of silverfollowed by analysis with stripping analysis.50

Enhancement by silver toward DNA detection is achieved by the silver particlesexhibiting a better electrochemical activity than gold, as the redox activity occursat lower potential (under 0.4V). Besides, silver metal can be easily oxidized tothe soluble silver metal ions in concentrated acid medium.86 This was shown byWang et al.50 where they reported a well-defined enhanced silver signal which was125 times greater than that of the gold-nanoparticle when silver is coated onto thegold-nanoparticle.

Gold and silver are not the only two metals that can be used as metal-nanoparticle labels. Zhu et al.87 reported an electrochemical stripping methodto detect DNA hybridization using cadmium sulfide. Their protocol consisted

June 23, 2007 12:9 WSPC/204-BRL 00047

182 E. L. S. Wong

Fig. 9. DNA detection scheme of metal-nanoparticle labeled electrochemical biosensor. The elec-trochemistry of metal-nanoparticle is observed (i) directly on the DNA-modified surface; (ii) afterdissolution of metal and (iii) the metal-nanoparticle is silver coated and the electrochemistry ofsilver is detected with or without silver dissolution. M = metal-nanoparticle.

of hybridization of the target DNA with the cadmium sulfide nanocluster DNAprobe, followed by anodic stripping analysis of the dissolved cadmium ions at amercury-coated glassy carbon electrode. In yet another innovative application ofthis approach, Willner et al.88 developed nanoparticle architectures of CdS particles

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 183

Fig. 10. Assembly of layered CdS-DNA aggregate.

and DNA to enhance the photoelectrochemical current that was used to detectDNA hybridization. The CdS-DNA aggregate was networked by repetitive treat-ment of CdS-modified DNA and complementary bridging DNA strands to producelarger assemblies of the CdS labels (Fig. 10). Therefore the large CdS-DNA net-work enhanced and amplified the photoelectrochemical current between the CdS-nanoparticle aggregate and the gold electrode.

Another attractive feature of employing metal-nanoparticles as labels is theability to detect multiple target DNA sequences simultaneously.89 Different targetDNA sequences can be encoded with different metal-nanoparticles to differentiatethe signals of the DNA targets obtained from stripping analysis of heavy metaldissolution. In one example, three different DNA targets were labeled with zincsulfide, cadmium sulfide and lead sulfide and after metal dissolutions, three well-defined and resolved stripping waves were observed at −1.12V (Zn), −0.68V (Cd)and −0.53 V (Pb) versus Ag|AgCl at the mercury-coated glassy carbon electrode.89

The potential and the current size of the individual peak revealed the identity andamount of the corresponding DNA target.

The metal-nanoparticle labeled approach is well-suited to multiple-target detec-tion. However, many development steps are involved in the assay of hybridization.Therefore, the robustness and reliability of the sensing interface can be problematic.The target samples will also be destroyed if anodic stripping analysis is performedas the dissolution of metal-nanoparticles is performed in acidic medium.

June 23, 2007 12:9 WSPC/204-BRL 00047

184 E. L. S. Wong

Fig. 11. Ferrocenenaphthalene diimide threading intercalator.

4.5. Intercalator labeled electrochemical DNA biosensors

A redox reporter molecule which gives greater information regarding whether aperfectly matched duplex has formed was shown by Takenaka et al.,46 who demon-strated that single-base mismatches could be detected electrochemically usinga threading intercalator ferrocenylnaphthalene diimide (Fig. 11). This threadingintercalator has twin benefits of not only binding four times more efficiently to theduplex over ss-DNA but the complexation occurs 80 times faster with the duplex.An anodic current at 520mV versus Ag|AgCl due to the electrochemistry of the fer-rocene derivatives was measured to detect hybridization. A current observed beforehybridization (i0) and after hybridization (i) was compared, and the percentageincrease [∆i = 100(i − i0)/i0] due to formation of DNA duplexes on the surface ofthe transducer was taken as the measure of the presence (and the amount) of thetarget DNA.

One strategy that has recently attracted attention, which requires no modifi-cation of target DNA and relies on the formation of a perfect DNA duplex fortransduction of DNA, is the concept of long-range electron transfer (Fig. 12). Theattractive feature of long-range electron transfer is that the transduction of DNAhybridization relies on charge transport through the perfect duplex via bondingbetween redox-active intercalator and the DNA base pair. The long-range electrontransfer approach is simple, requires no labeling of target DNA. This approach isuniquely well-suited for mismatch detection as any perturbation of the base pairin the DNA duplex will affect the efficiency of the rate of electron transfer andtherefore an attenuation of electrochemical signal will be observed.

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 185

Fig. 12. Detection of DNA hybridization via long-range electron transfer.

Long-range electron transfer for detecting DNA hybridization has been demon-strated electrochemically by Barton and co-workers.14,45,90,91 In these electrochem-ical systems, the DNA duplexes are assembled into densely packed films on a goldelectrode surface. Long-range electron transfer is demonstrated via the electrochem-istry of electroactive unbound intercalators methylene blue (MB)14,45,91 or tethereddaunomycin90 which intercalates into the end of the DNA duplexes remote from theelectrode. The rate of electron transfer for the 18 base pairs with the daunomycinintercalated at the end was observed to be 100 s−1 with little attenuation in ratewith distance.14 It is important to note that, unlike the photoinduced electron trans-fer studies, ground state electron transfer is achieved in the electrochemical case.92

Denaturing the duplex to leave only single strands on the electrode surface resultedin a diminution in the charge passed where some charge is observed with the ss-DNAmodified electrodes as the denaturing of the closely packed duplex leaves a moreopen interface and the redox species can access the electrode directly.45 The impor-tance of the intercalation of the redox molecule in enabling efficient electron transferwas illustrated by Heller and co-workers93 where a similar densely packed DNA film

June 23, 2007 12:9 WSPC/204-BRL 00047

186 E. L. S. Wong

was prepared. The redox active molecule, pyrrolo–quinoline–quinone (PQQ) wascovalently attached to the end of the DNA duplexes but with a sufficiently shorttail to prevent intercalation. With the PQQ essentially locating on the top of the12 base pair duplexes the rate of electron transfer was significantly slower (1.5 s−1).Heller and co-workers hypothesized that the highly ordered DNA films could beregarded as “ionic crystals” which allowed free movement of electrons between thePQQ and the electrode.

Gooding and co-workers further extended the concept of long-range electrontransfer for DNA transduction by using an anionic intercalator, anthraquinone-2,6-disulfonic acid (AQDS) and a loosely-packed ss-DNA film.15 The ability of thisbiosensing system to differentiate a complementary DNA sequence from a non-complementary target DNA sequence and one containing a single-base pair mis-matches (including thermodynamically stable G–A mismatch target) is consistentwith Barton’s group.14,45,90,91 The absence of background electrochemical currentwhen on single strand probe DNA was present on the interface was attributed tothe electrostatic repulsion between the AQDS and the sensing interface preventingany non-specific binding of AQDS to the electrode surface. The lack of backgroundelectrochemical current then allows any current to be attributed to long-range elec-tron and has the advantage from a sensing perspective that not all probe strandsare required to have formed duplexes to give a reliable signal. However, the DNAbiosensor based on long-range electron transfer using AQDS has limitations withregard to high detection limit and a long assay time. Nevertheless, these limitationswere overcome by using a less anionic intercalator, anthraquinone-2-monosulfonicacid (AQMS)94 and a more favorable DNA interface.95 The newly developed in situapproach by Gooding group,96 where the DNA transduction was performed in asingle step with AQMS present in the sample solution containing the target DNA,also allows DNA hybridization to be monitored in real time with good selectivityand sensitivity.

5. Future Prospects

The development of DNA hybridization biosensors is already quite vast and con-tinues to grow and broaden. Successful detection of DNA hybridization and DNAmutation has been demonstrated using different types of DNA hybridization sen-sors as discussed above. DNA microarrays or DNA chips are commercially avail-able exploiting different fabrication technologies and detection protocols. Theseinclude (i) optical-based DNA Chips as developed by Affymetrix Inc. (GeneChip),97 Protogene laboratories, Hyseq Inc. (HyChipTM) and Nanogen98; (ii) massspectrometry-based arrays as developed by Sequenom and (iii) electrochemical-based DNA chips as developed by GeneOhm Sciences Inc., Toshiba Corp. andMotorola Life Sciences Inc. Despite the promising opportunities offered by theseDNA sensing technologies for large-scale genomic diagnostics, the DNA assays rou-tinely start with amplification by PCR. Without the preamplification step, low copy

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 187

numbers of DNA sample which gives rise to challenges for interfacial hybridization.Therefore, the real goal is to eliminate the need for PCR and develop a rapid assayto detect a few copies of DNA on a DNA chip. With the rapid advances in microfab-rication and biosensing world, it can be foreseen that this goal will be accomplishedin the near future.

References

1. J. Wang, Nucleic Acids Res. 28, 3011 (2000).2. G. H. Keller and M. M. Manak, DNA Probes: Molecular Hybridization Technology

(Stockton Press, New York, 1993).3. J. Fan, D. Wang, A. Berno, C. J. Siao, G. Ghandour, L. Hsie, P. Young, N. Perkins,

J. Spencer, D. Morris, L. Kruglyak, R. Sapolsky, C. Nusbaum, T. Topaloglou,T. Hawkins, L. Stein, S. Rozen, R. Lipshutz, T. Hudson, M. Chee and E. S. Lan-der, Am. J. Hum. Genet. 61, A275 (1997).

4. P. Nilsson, B. Persson, M. Uhlen and P. A. Nygren, Anal. Biochem. 224, 400 (1995).5. P. A. E. Piunno, U. J. Krull, R. H. E. Hudson, M. J. Damha and H. Cohen, Anal.

Chem. 67, 2635 (1995).6. A. P. Abel, M. G. Weller, G. L. Duveneck, M. Ehrat and H. M. Widmer, Anal. Chem.

68, 2905 (1996).7. S. Pilevar, C. C. Davis and F. Portugal, Anal. Chem. 70, 2031 (1998).8. X. H. Fang, J. W. J. Li, J. Perlette, W. H. Tan and K. M. Wang, Anal. Chem. 72,

747a (2000).9. X. H. Fang, X. J. Liu, S. Schuster and W. H. Tan, J. Am. Chem. Soc. 121, 2921 (1999).

10. Y. Okahata, T. Kobayashi, K. Tanaka and M. Shimomura, J. Am. Chem. Soc. 120,6165 (1998).

11. A. K. Boal and J. K. Barton, Bioconjugate Chem. 16, 312 (2005).12. C. Xu, P. G. He and Y. Z. Fang, Anal. Chim. Acta, 411, 31 (2000).13. K. M. Millan and S. R. Mikkelsen, Anal. Chem. 65, 2317 (1993).14. S. O. Kelley, J. K. Barton, N. M. Jackson and M. G. Hill, Bioconjugate Chem. 8, 31

(1997).15. E. L. S. Wong and J. J. Gooding, Anal. Chem. 75, 3845 (2003).16. C. Xu, H. Cai, Q. Xu, P. G. He and Y. Z. Fang, Fres. J. Anal. Chem. 369, 428 (2001).17. E. S. Yeung and S. H. Kang, Abstr. Pap. Am. Chem. Soc. 222, U378 (2001).18. V. Chan, D. J. Graves, P. Fortina and S. E. McKenzie, Langmuir 13, 320 (1997).19. Y. Belosludtsev, B. Iverson, S. Lemeshko, R. Eggers, R. Wiese, S. Lee, T. Powdrill

and M. Hogan, Anal. Biochem. 292, 250 (2001).20. S. V. Lemeshko, T. Powdrill, Y. Y. Belosludtsev and M. Hogan, Nucleic Acids Res.

29, 3051 (2001).21. P. Caillat, D. David, M. Belleville, F. Clerc, C. Massit, F. Revol-Cavalier, P. Peltie,

T. Livache, G. Bidan, A. Roget and E. Crapez, Sens. Actuators B, 61, 154 (1999).22. T. Livache, A. Roget, E. Dejean, C. Barthet, G. Bidan and R. Teoule, Nucleic Acids

Res. 22, 2915 (1994).23. R. Levicky, T. M. Herne, M. J. Tarlov and S. K. Satija, J. Am. Chem. Soc. 120, 9787

(1998).24. E. Huang, F. M. Zhou and L. Deng, Langmuir 16, 3272 (2000).25. J. A. Ferguson, T. C. Boles, C. P. Adams and D. R. Walt, Nat. Biotechnol. 14, 1681

(1996).26. H. J. Watts, D. Yeung and H. Parkes, Anal. Chem. 67, 4283 (1995).27. R. Georgiadis, K. P. Peterlinz and A. W. Peterson, J. Am. Chem. Soc. 122, 3166 (2000).

June 23, 2007 12:9 WSPC/204-BRL 00047

188 E. L. S. Wong

28. J. J. Storhoff, R. Elghanian, R. C. Mucic, C. A. Mirkin and R. L. Letsinger, J. Am.Chem. Soc. 120, 1959 (1998).

29. T. A. Taton, C. A. Mirkin and R. L. Letsinger, Science 289, 1757 (2000).30. S. Yamaguchi, T. Shimomura, T. Tatsuma and N.Oyama, Anal. Chem. 65, 1925 (1993).31. X. D. Su, R. Robelek, Y. J. Wu, G. Y. Wang and W. Knoll, Anal. Chem. 76, 489

(2004).32. Y. Okahata, Y. Matsunobu, K. Ijiro, M. Mukae, A. Murakami and K. Makino, J. Am.

Chem. Soc. 114, 8299 (1992).33. K. Ito, K. Hashimoto and Y. Ishimori, Anal. Chim. Acta, 327, 29 (1996).34. N. C. Fawcett, J. A. Evans, L. C. Chien and N. Flowers, Anal. Lett. 21, 1099 (1988).35. G. Sauerbrey, Z. Phys. 155, 206 (1959).36. J. Wang, P. E. Nielsen, M. Jiang, X. H. Cai, J. R. Fernandes, D. H. Grant, M. Ozsoz,

A. Beglieter and M. Mowat, Anal. Chem. 69, 5200 (1997).37. A. Bardea, A. Dagan, I. Ben-Dov, B. Amit and I. Willner, Chem. Commun. 7, 839

(1998).38. F. Caruso, E. Rodda, D. F. Furlong, K. Niikura and Y. Okahata, Anal. Chem. 69,

2043 (1997).39. K. Niikura, K. Nagata and Y. Okahata, Chem. Lett. 25, 863 (1996).40. H. B. Su, P. Williams and M. Thompson, Anal. Chem. 67, 1010 (1995).41. K. Kerman, M. Kobayashi and E. Tamiya, Meas. Sci. Technol. 15, R1 (2004).42. T. G. Drummond, M. G. Hill and J. K. Barton, Nat. Biotechnol. 21, 1192 (2003).43. J. Wang, Anal. Chim. Acta 469, 63 (2002).44. J. J. Gooding, Electroanalysis 14, 1149 (2002).45. S. O. Kelley, E. M. Boon, J. K. Barton, N. M. Jackson and M. G. Hill, Nucleic Acids

Res. 27, 4830 (1999).46. S. Takenaka, K. Yamashita, M. Takagi, Y. Uto and H. Kondo, Anal. Chem. 72, 1334

(2000).47. T. deLumleyWoodyear, C. N. Campbell and A. Heller, J. Am. Chem. Soc. 118, 5504

(1996).48. D. J. Caruana and A. Heller, J. Am. Chem. Soc. 121, 769 (1999).49. C. N. Campbell, D. Gal, N. Cristler, C. Banditrat and A. Heller, Anal. Chem. 74,

158 (2002).50. J. Wang, R. Polsky and D. K. Xu, Langmuir 17, 5739 (2001).51. J. Wang, D. K. Xu and R. Polsky, J. Am. Chem. Soc. 124, 4208 (2002).52. M. Ozsoz, A. Erdem, K. Kerman, D. Ozkan, B. Tugrul, N. Topcuoglu, H. Ekren and

M. Taylan, Anal. Chem. 75, 2181 (2003).53. F. Lucarelli, G. Marrazza, I. Palchetti, S. Cesaretti and M. Mascini, Anal. Chim. Acta

469, 93 (2002).54. D. Ozkan, A. Erdem, P. Kara, K. Kerman, B. Meric, J. Hassmann and M. Ozsoz,

Anal. Chem. 74, 5931 (2002).55. E. Palecek, S. Billova, L. Havran, R. Kizek, A. Miculkova and F. Jelen, Talanta 56,

919 (2002).56. J. Wang, G. Rivas, J. R. Fernandes, J. L. L. Paz, M. Jiang and R. Waymire, Anal.

Chim. Acta 375, 197 (1998).57. J. J. Gooding, A. Chou, F. J. Mearns, E. Wong and K. L. Jericho, Chem. Commun.

1938 (2003).58. E. Palecek, Nature 188, 656 (1960).59. F. Jelen, M. Fojta and E. Palecek, J. Electroanal. Chem. 427, 49 (1997).60. J. Wang, X. H. Cai, B. M. Tian and H. Shiraishi, Analyst 121, 965 (1996).61. S. C. Casegreen and E. M. Southern, Nucleic Acids Res. 22, 131 (1994).

June 23, 2007 12:9 WSPC/204-BRL 00047

Recent Trends in DNA Biosensing Technologies 189

62. H. H. Thorp, Trends Biotechnol. 16, 117 (1998).63. J. Wang, X. H. Cai, G. Rivas and H. Shiraishi, Anal. Chim. Acta 326, 141 (1996).64. J. Wang and A. B. Kawde, Analyst 127, 383 (2002).65. H. Shiraishi and R. Takahashi, Bioelectrochem. Bioenerg. 31, 203 (1993).66. J. Wang, A. N. Kawde and M. Musameh, Analyst 128, 912 (2003).67. K. B. Wu, J. J. Fei, W. Bai and S. S. Hu, Anal. Bioanal. Chem. 376, 205 (2003).68. K. Kerman, Y. Morita, Y. Takamura, M. Ozsoz and E. Tamiya, Electroanalysis 16,

1667 (2004).69. M. L. Pedano and G. A. Rivas, Electrochem. Commun. 6, 10 (2004).70. H. W. Fink and C. Schonenberger, Nature 398, 407 (1999).71. Y. T. Long, C. Z. Li, T. C. Sutherland, H. B. Kraatz and J. S. Lee, Anal. Chem. 76,

4059 (2004).72. J. S. Lee, L. J. P. Latimer and R. S. Reid, Biochem. Cell. Biol. 71, 162 (1993).73. H. Korri-Youssoufi, F. Garnier, P. Srivastava, P. Godillot and A. Yassar, J. Am. Chem.

Soc. 119, 7388 (1997).74. A. Erdem, K. Kerman, B. Meric, U. S. Akarca and M. Ozsoz, Anal. Chim. Acta 422,

139 (2000).75. A. Erdem, K. Kerman, B. Meric and M. Ozsoz, Electroanalysis 13, 219 (2001).76. W. R. Yang, M. Ozsoz, D. B. Hibbert and J. J. Gooding, Electroanalysis 14, 1299

(2002).77. C. H. Fan, K. W. Plaxco and A. J. Heeger, Proc. Natl. Acad. Sci. USA 100, 9134

(2003).78. R. M. Umek, S. W. Lin, J. Vielmetter, R. H. Terbrueggen, B. Irvine, C. J. Yu, J. F.

Kayyem, H. Yowanto, G. F. Blackburn, D. H. Farkas, and Y. P. Chen, J. Mol. Diagn.3, 74 (2001).

79. S. D. Vernon, D. H. Farkas, E. R. Unger, V. Chan, D. L. Miller, Y. P. Chen, G. F.Blackburn and W. C. Reeves, BMC Infect. Dis. 3 (2003).

80. M. Dequaire and A. Heller, Anal. Chem. 74, 4370 (2002).81. F. Patolsky, M. Zayats, E. Katz and I. Willner, Anal. Chem. 71, 3171 (1999).82. L. Alfonta, A. K. Singh and I. Willner, Anal. Chem. 73, 91 (2001).83. F. Patolsky, A. Lichtenstein and I. Willner, Nat. Biotechnol. 19, 253 (2001).84. F. Patolsky, A. Lichtenstein and I. Willner, Angew. Chem. Int. Ed. 39, 940 (2000).85. J. Wang, D. K. Xu, A. N. Kawde and R. Polsky, Anal. Chem. 73, 5576 (2001).86. H. Cai, Y. Xu, N. N. Zhu, P. G. He and Y. Z. Fang, Analyst 127, 803 (2002).87. N. N. Zhu, A. P. Zhang, P. G. He and Y. Z. Fang, Analyst 128, 260 (2003).88. I. Willner and B. Willner, Pure Appl. Chem. 74, 1773 (2002).89. J. Wang, G. D. Liu and A. Merkoci, J. Am. Chem. Soc. 125, 3214 (2003).90. S. O. Kelley, N. M. Jackson, M. G. Hill and J. K. Barton, Angew. Chem. Int. ed. 38,

941 (1999).91. E. M. Boon, D. M. Ceres, T. G. Drummond, M. G. Hill and J. K. Barton, Nat.

Biotechnol. 18, 1096 (2000).92. N. M. Jackson and M. G. Hill, Curr. Opin. Chem. Biol. 5, 209 (2001).93. G. Hartwich, D. J. Caruana, T. de Lumley-Woodyear, Y. B. Wu, C. N. Campbell and

A. Heller, J. Am. Chem. Soc. 121, 10803 (1999).94. E. L. S. Wong, P. Erohkin and J. J. Gooding, Electrochem. Commun. 6, 648 (2004).95. E. L. S. Wong, E. Chow and J. J. Gooding, Langmuir 21, 6957 (2005).96. E. L. S. Wong and J. J. Gooding, Anal. Chem. 78, 2138 (2006).97. A. C. Pease, D. Solas, E. J. Sullivan, M. T. Cronin, C. P. Holmes, and S. P. A. Fodor,

Proc. Natl. Acad. Sci. USA, 91, 5022 (1994).98. C. F. Edman, D. E. Raymond, D. J. Wu, E. G. Tu, R. G. Sosnowski, W. F. Butler,

M. Nerenberg and M. J. Heller, Nucleic Acids Res. 25, 4907 (1997).