Maximizing Mismatch Discrimination by Surface-Tethered Locked Nucleic Acid Probes via Ionic Tuning

Maximizing Mismatch Discrimination by Surface-Tethered LockedNucleic Acid Probes via Ionic TuningSourav Mishra, Srabani Ghosh, and Rupa Mukhopadhyay*

Department of Biological Chemistry, Indian Association for the Cultivation of Science, Jadavpur, Kolkata-700032, India

*S Supporting Information

ABSTRACT: Several investigations on DNA-based nucleicacid sensors performed in the past few years point toward therequirement of an alternative nucleic acid that can detect targetDNA strands more efficiently, i.e., with higher sensitivity andselectivity, and can be more robust compared to the DNAsensor probes. Locked nucleic acid (LNA), a conformationallyrestricted DNA analogue, is potentially a better alternativethan DNA, since it is nuclease-resistant, it can form a morestable duplex with DNA in a sequence-specific manner, and itinteracts less with substrate surface due to presence of a rigidbackbone. In this work, we probed solid-phase dehybridization of ssDNA targets from densely packed fully modified ssLNAprobes immobilized onto a gold(111) surface by fluorescence-based measurement of the “on-surface” melting temperatures. Wefind that mismatch discrimination can be clearly improved by applying the surface-tethered LNA probes, in comparison to thecorresponding DNA probes. We show that concentration as well as type of cation (monovalent and polyvalent) can significantlyinfluence thermal stability of the surface-confined LNA−DNA duplexes, the nature of concentration dependence contradictingthe solution phase behavior. Since the ionic setting influenced the fully matched duplexes more strongly than the singlymismatched duplexes, the mismatch discrimination ability of the surface-confined LNA probes could be controlled by ionicmodulations. To our knowledge, this is the first report on ionic regulation of melting behavior of surface-confined LNA−DNAduplexes.

Identification of single base mismatch in DNA sequences is ofgreat importance to understand genetic variations present

among individuals; to know how the individuals developresponse to external agents like drugs, pathogens, chemicalsetc.; and to recognize an individual’s propensity towarddevelopment of a specific disease. There are around 10 millionsingle nucleotide polymorphisms (SNPs) that have beendetected in the human genome.1 Generally identification ofthese SNPs relies upon detection of the differential responsebetween hybridization of the fully matched sequences andsingly mismatched sequences or between dehybridization of thefully matched duplexes and singly mismatched duplexes. TheSNPs are usually detected/analyzed by real-time polymerasechain reaction (PCR),2−5 microarrays,6,7 and nanobiosensortechnologies.8,9 A significant number of approaches that arebased on microarray and biosensor technologies may requiredetection of solid-phase hybridization, in which surface-immobilized capture probes bind to target molecules fromsolution. In the past two decades, these technologies have madeconsiderable advancements and have been applied in importantareas like gene expression profiling, genotyping, and biologicaldetection.10−12 Though the underlying facts behind nucleic acidhybridization in bulk solution are fairly well-understood, theunderstanding of the nucleic acid hybridization onto solidsurfaces is much less developed.10,13 Recently, there have beensome reports on DNA-biosensors that are based on solid-phaseDNA−DNA hybridization.8,10,14−18 In these reports, the effects

of changes in various parameters like oligonucleotide length,14

salt concentration,14 type of cations,14 and type of buffers18

have been monitored to understand their role in stability of thesurface-confined DNA−DNA duplexes as well as to find out theoptimum situation to maximize single base mismatchdiscrimination.14 Though DNA-based biosensors have foundwide applications in the microscale9 as well as nanoscale nucleicacid sensing experiments,19−22 its reduced bioactivity due topotential DNA−surface interactions through relatively exposednucleobases and degradability by the nuclease compel one tosearch for alternatives so that these drawbacks can beovercome. Recently, it has been shown that locked nucleicacid (LNA) probes can potentially be a better alternative thanthe DNA probes for single base mismatch discrimination intarget DNA sequences onto a gold(111) surface.23 In thepresent study, we explored an approach for solid-state singlebase mismatch discrimination in target DNA sequences usingthe LNA sensor probes, based on measurement of the “on-surface” melting temperature (Tm) values of the LNA−DNAand the DNA−DNA duplexes. This was done by means offluorescence detection of the Cy3 labeled target DNA probesthat remained on the surface at the end of each heat-induced

Received: October 3, 2012Accepted: December 26, 2012Published: December 26, 2012

Article

pubs.acs.org/ac

© 2012 American Chemical Society 1615 dx.doi.org/10.1021/ac3028382 | Anal. Chem. 2013, 85, 1615−1623

pubs.acs.org/ac

dehybridization step. The sensitivity of this approach wasestimated from the differences between the melting temper-atures of the respective duplexes and maximized by controllingthe ionic parameters like salt concentration and type of cation.LNA is a conformationally restricted molecule since it

contains a modified ribose moiety in which the 2′-oxygen andthe 4′-carbon are linked by a methylene bridge, in effect,locking the sugar in a RNA mimicking sugar conformation (N-type) (Figure 1).24,25 LNA can bind with complementary

DNA/RNA sequences in a sequence-specific manner obeyingthe Watson−Crick base pairing rule with higher affinitycompared to DNA, which is reflected in the higher values ofthe solution melting temperatures of LNA−DNA/LNA−RNAduplexes compared to those of DNA−DNA/DNA−RNAduplexes.24,25 LNA is nuclease-resistant;26,27 its higher struc-tural rigidity may prevent interactions with the solidsubstrates,28 and it can have multiple water bridges thatprovide it with extra stability compared to DNA or RNA.29

Though it was found that LNA can enhance single basemismatch discrimination in both bulk solutions30 as well as onthe surface,23 the basic framework of hybridization “in solutionphase” and “on surface” can differ significantly with respect tovarious aspects as following. Since the probe density and localconcentration of the negatively charged oligonucleotides couldbe considerably higher on the surface compared to that in bulksolution, both steric effects as well as electrostatic effects couldhave an important role in solid-phase hybridization.14 Ingeneral, cations compensate the negative charge of theoligonucleotide backbone and therefore stabilize the oligonu-cleotide duplexes.14 So, it is expected that the effect of cationswill be more pronounced in solid-phase hybridizationcompared to hybridization processes in bulk solution becauseof high surface density of negatively charged probes onto thesolid surface. It has been reported that hybridization efficiencycould be profoundly influenced by the total concentration ofsodium ion irrespective of the type of buffer used.18 Themajority of the investigations so far dealing with the effect ofcations on solid-phase nucleic acid hybridization have focusedmore on probe density, while the effect of different types andconcentrations of the cations, present in the hybridizationbuffer, on duplex stability has been investigated to a little

extent. Very recently, Springer et al.14 addressed these facts, i.e.,the effect of cations (both the type and the concentration) onthe duplex stability, mismatch discrimination for solid-phaseDNA−DNA hybridization to an extent but the effect of theseparameters on solid phase LNA-DNA hybridization, with anaim to maximize single base mismatch discrimination, is yet tobe addressed.Herein, for the first time, we have investigated the effects of

varying concentrations of monovalent and polyvalent cations,namely Na+, Mg2+, spermidine (3+), and spermine (4+) onsolid-phase hybridization of fully matched and singlymismatched DNA targets to densely packed immobilizedLNA probes onto a gold(111) surface. We found that in case ofboth LNA and DNA probes, single base mismatch discrim-ination is better performed on the surface and that LNA excelsDNA in this respect. Importantly, the type as well asconcentration of the cations is found to influence the mismatchdiscrimination ability of the LNA probes.

■ MATERIALS AND METHODSPreparation of LNA Sensor Probe Solutions. The

solutions of thiolated and fully modified LNA sequences (Table1) (HPLC purified, procured from Exiqon, Denmark) were

prepared in phosphate buffered saline (20 mM sodiumphosphate, 100 mM sodium chloride, pH 7.0) or PBS. TheLNA concentrations were determined by UV−visible spectros-copy, considering the ε260 (L/(mol × cm)) values for LNA-1,LNA-2, and LNA-3 as 112 700, 118 100, and 111 100,respectively (all the ε260 values presented here or later wereobtained from the manufacturer-provided data sheets). LNA-1and LNA-2 were the sensor probes, LNA-1 for a completematch situation and LNA-2 for a single base mismatchsituation. LNA-3 was the fully mismatched sensor probe usedfor control experiments.

Preparation of DNA Sensor Probe Solutions. Thesolutions of thiolated DNA sequences (Table 1) (HPLCpurified, procured from alpha DNA, Canada) were prepared inthe PBS medium. The DNA concentrations were determinedby UV−visible spectroscopy, considering the ε260 (L/(mol ×cm)) values for DNA-1, DNA-2, and DNA-3 as 123 020, 131350, and 123 020, respectively. DNA-1 and DNA-2 were thesensor probes, DNA-1 for complete match situation and DNA-2 for a single base mismatch situation. DNA-3 was the fullymismatched sensor probe used for control experiments.

Preparation of Cy3 Labeled Target DNA Solutions.The solutions of Cy3 labeled DNA sequences (Table 1)(HPLC purified, procured from IDT, Canada) were preparedin the sodium phosphate buffer (20 mM Na2HPO4/20 mMNaH2PO4, pH 7.0) containing a selected amount of 50, 100,

Figure 1. Chemical structures of (A) deoxyribonucleic acid (DNA)and (B) locked nucleic acid (LNA).

Table 1. Nucleic Acid Sequences Used in the Present Study

DNA/LNA sequence

DNA-1 5′-HS-C6-CTA-TGT-CAG-CAC-3′DNA-2 5′-HS-C6-CTA-TGT-AAG-CAC-3′DNA-3 5′-HS-C6-CGA-TCT-GCT-AAC-3′LNA-1 5′-HS-C6-CTA-TGT-CAG-CAC-3′LNA-2 5′-HS-C6-CTA-TGT-AAG-CAC-3′LNA-3 5′-HS-C6-CGA-TCT-GCT-AAC-3′Cy3 DNA-1 5′-Cy3-GTG-CTG-ACA-TAG-3′Cy3 DNAnc 5′-Cy3-CGA-TCT-GCT-AAC-3′T-DNA-1 5′- GTG-CTG-ACA-TAG-3′

Analytical Chemistry Article

dx.doi.org/10.1021/ac3028382 | Anal. Chem. 2013, 85, 1615−16231616

500, 1000, 2000, 3000 mM NaCl or 1.5, 5, 10, 15, 20 mMMgCl2, or 5, 10, 15 mM spermidine, or 5, 10 mM spermine.The DNA concentrations were determined by UV−visiblespectroscopy, considering the ε260 (L/(mol × cm)) value forCy3 DNA-1 and Cy3 DNAnc as 124 400 and 123 200,respectively. Cy3 DNA-1 was the labeled target probe, beingfully matched to LNA-1 and DNA-1, and bearing a single basemismatch with respect to both LNA-2 and DNA-2. Cy3 DNAncwas the fully mismatched target probe used for controlexperiments.Preparation of Unlabeled Target DNA Solution. The

solution of unlabeled target DNA sequence T-DNA-1 (Table1) (HPLC purified, procured from alpha DNA, Canada) wasprepared in the PBS medium and its concentration wasdetermined by UV−visible spectroscopy considering the ε260(L/(mol × cm)) value as 133 300. T-DNA-1 was fully matchedto LNA-1 and DNA-1 and singly mismatched to LNA-2 andDNA-2.Preparation of Gold(111) Surface. Gold on mica (Phasis,

Switzerland) substrate (gold layer thickness, 200 nm) was flameannealed until a reddish glow appeared. This procedure wasrepeated 7−8 times. After a short period (1−2 s) of cooling inair, the substrate was subjected to further modification steps.Preparation of DNA/LNA Sensor Probe Modified

Gold(111) Surface. In all the cases, immobilization of thenucleic acid sensor probes on gold(111) surface was performedin the PBS medium at room temperature (24 ± 1 °C).Hybridization was carried out in 20 mM sodium phosphatebuffer (20 mM Na2HPO4, 20 mM NaH2PO4, pH 7.0)containing a selected amount of 50, 100, 500, 1000, 2000,3000 mM NaCl, or 1.5, 5.0, 10, 15, 20 mM MgCl2, or 5, 10, 15mM spermidine, or 5, 10 mM spermine at room temperature.Fluorescence Intensity Measurement and Determi-

nation of Melting Temperature (Tm) on Surface. Freshlyannealed gold(111) substrate was immersed in thiolated LNAsolution of 0.5 μM concentration and incubated for 4 h at roomtemperature. After incubation, the substrate was first washedwith 1 mL (2 × 500 μL) of PBS solution followed by 2 mL (4× 500 μL) of filtered autoclaved Milli-Q water to remove thenonspecifically adsorbed molecules and then dried under gentlenitrogen jet. Then, the LNA-modified gold pieces weresubjected to hybridization by incubating into 1 μM Cy3DNA-1 solution, having different hybridization environment,i.e., hybridization buffer containing a selected amount of 50,100, 500, 1000, 2000, 3000 mM NaCl, or 1.5, 5, 10, 15, 20 mMMgCl2, or 5, 10, 15 mM spermidine, or 5, 10 mM spermine asdesired, for 1 h at room temperature. Next, the gold pieces weretaken out from the hybridization solution, washed thoroughlywith 1 mL (2 × 500 μL) of corresponding hybridization bufferfollowed by 2 mL (4 × 500 μL) of filtered autoclaved Milli-Qwater to remove the nonspecifically bound target molecules anddried under a mild nitrogen jet. The fluorescence images wereobtained at room temperature in a dark condition with anOlympus BX61 fluorescence microscope by excitation (λexc) at∼550 nm and emission (λem) at ∼570 nm. Then, to monitorthe thermal denaturation behavior of the surface-confinedLNA−DNA duplexes, the gold piece was placed into 600 μL of20 mM sodium phosphate buffer (20 mM sodium phosphate,100 mM sodium chloride, pH 7.0) and heated to a desiredtemperature for 15 min. Then, the gold piece was taken out,washed and fluorescence images were recorded again in orderto find out the amount of Cy3-labeled target molecules left onthe surface. This process was continued until all the Cy3-

labeled target molecules were removed from the surface, asconfirmed by fluorescence imaging (see Figures S1 and S2 inthe Supporting Information). Heating was performed in severalsteps till the complete removal of the Cy3-labeled targets fromthe surface was ensured. Exactly the same protocol was appliedto monitor the thermal denaturation behavior of the surface-confined DNA−DNA duplexes by taking thiolated DNAoligonucleotides as sensor molecules. The exposure time waskept fixed for all the fluorescence imaging experiments. Thefluorescence intensity corresponding to every respective imagewas determined by Image-pro MC6.1 software (MediaCybernetics, Bethesda, MD). Then, the decrease in fluores-cence intensity with an increase in temperature was plotted. Tocalculate the melting temperature from the experimental data, asigmoidal fit was accomplished employing the Boltzmanfunction using the data evaluation software Origin8 (OriginLabCooperation, Northampton, MA). The equation used for fittingwas y = A2 + (A1 − A2)/(1 + exp((x − x0)/dx)), where A1 =initial y value, A2 = final y value, and x0 = center, i.e., the valueof x at (A1 + A2)/2, dx = time constant where the constraint isdx! = 0. The melting temperatures were calculated from theinflection point of the fit function as described earlier.31 Thestandard error of melting temperature measurement was ±0.2°C.

Determination of Tm in Solution. The oligonucleotidesolutions were prepared in 20 mM sodium phosphate buffer(pH 7.0) containing a selected amount of 50, 100, 500 mMNaCl. Hybridization was performed by mixing equal volumes ofequimolar solutions of target DNA and the relevant LNA/DNAsensor probe sequences at room temperature and allowing theresulting solution to stand for 30 min. Then, the absorbance at260 nm was measured with Peltier control Perkin-Elmer DTP1UV−vis spectrophotometer. The heating cycles were per-formed considering the temperature range of 25−96 °C at aheating rate of 1 °C/min. In order to calculate the meltingtemperatures, the fraction of melted base pairs, θ, wascalculated from the standard formula, θ = (A − AL)/(AU −AL), where A, AL, and AU are sample absorbance, absorbance ofthe lower baseline, and absorbance of the upper baseline,respectively. Tm is defined as the temperature where θ = 0.5.32

To check the reversibility of melting transitions, cooling curveswere collected for fully matched and singly mismatched LNA−DNA duplexes for the salt concentration of 100 mM at anannealing rate of 1 °C/min.

■ RESULTS AND DISCUSSIONIn this study, the single base mismatch discrimination capabilityof fully modified thiol-LNA probes immobilized ontogold(111) surface has been maximized by ionic tuning, i.e.,by varying salt concentration and the type of cations present inhybridization buffer. The applied condition for formation of theself-assembled LNA sensor films on a gold(111) surface wasoptimized before for achieving high coverage and a molecularorientation away from the surface.23 The sensor strandmodified gold samples were always prepared by the immersionmethod so that the nucleic acid molecules could be kept inwell-solvated conditions during the preparative stage. Effective-ness of the immersion method over the other samplepreparation methods, e.g., droplet contact method and dropletdeposition method, was exemplified earlier.33 Gold(111)substrate was selected since this surface is widely used inbiosensor applications,19,20,34,35 especially where immobiliza-tion of the sensor molecules via gold−thiol linkage formation36



is relevant. Since cleanliness of the substrate surface is anabsolute necessity for effective anchoring of the thiolatednucleic acid probes on the gold surface,36 the gold pieces werealways freshly annealed prior to nucleic acid modification.Three 12-mer thiolated ssLNA sequences, namely, LNA-1,LNA-2, and LNA-3, were employed as the sensor probes(Table 1). The 12-mer 5′-Cy3 labeled ssDNA sequence (Cy3DNA-1), which was fully matched to LNA-1 and singlymismatched to LNA-2, was employed as the target strand(Table 1). The Cy3 DNAnc was the fully mismatched targetprobe used for control experiments (Table 1). Three DNAsequences DNA-1, DNA-2, and DNA-3 having the same basesequence as LNA-1, LNA-2, and LNA-3, respectively, wereapplied as the DNA sensor probes (Table 1). The sensorprobes were kept small in size, and the mismatch in both LNAand DNA sensor probes was centrally placed in order tomaximize mismatch discrimination capacity of the probes. In allthe sensor probe molecules, a hexyl spacer [−(CH2)6−] wasintroduced at the 5′-end, which helps to keep the nucleic acidpart away from the gold surface so that nonspecific adsorptionvia nucleobases could be avoided to a considerable extent. TheLNA sequence taken was fully LNA-modified since Owczarzyet al. reported very recently that additional LNA moietiesadjacent to the initial modification seem to enhance stackingand H-bonding interactions,37 which is expected to result inhigh affinity to the complementary target DNA strands.Moreover, full modification could ensure the maximum levelof nuclease resistance of the sequence. Both the sensor and thetarget sequences were kept short in length (12 mer) since it hasbeen shown that duplex stabilization is achieved best with shortoligonucleotide sequences.14 LNA purine (adenine) was placedat the mismatch site in order to achieve the maximummismatch discrimination as per an earlier finding that LNApurines offer great potential to recognize mismatches than LNApyrimidines and DNA purines.38 Fluorescence microscopy wasemployed to determine the melting temperatures of thesurface-confined LNA−DNA and DNA−DNA duplexes,formed in different ionic environments, to identify theoptimum situation where the single base mismatch discrim-ination ability of the LNA film could be maximized. Thesolution phase melting temperatures of the sensor−targetduplexes, same as those used for on-surface measurements,were determined by UV−vis absorption spectroscopy.Tm Measurement of the Surface-Confined LNA−DNA

and DNA−DNA Duplexes. To investigate the meltingbehavior of the surface-bound LNA−DNA and DNA−DNAduplexes, effective anchoring of the thiolated LNA and DNAsensor probes onto the gold(111) surface was ensured first.The modified gold surfaces were then exposed to the labeledtarget DNA probes, and the fluorescence images were captured.To determine the Tm values, the gold pieces were heated todesired temperatures by starting from the lower temperaturesand in steps that could be as small as 0.7 °C (near theanticipated melting temperature value) or as high as 5.0 °C(away from the melting temperature value). Since a thoroughwash was given after each step of heating that should ensureeffective removal of the dehybridized Cy3 DNA-1 strands fromthe surface, the measure of the fluorescence intensity obtainedafter each heating step should be directly proportional to theremaining portion of the duplexes on the surface. With anincrease in temperature, fluorescence intensity correspondingto the labeled target molecules on the surface was found toreduce and it became almost negligible after reaching a

particular temperature for each type of duplex, quite obviouslybecause the duplexes were increasingly dehybridized with anincrease in temperature. The fluorescence intensity values wereplotted against temperature in each case, and from thedenaturation profiles the Tm values were determined (Figure2). The “on-surface” Tm values differed significantly from the

corresponding melting temperatures obtained in bulk solution,indicating largely different energetics playing on the surfacecompared to that in solution. Importantly, single base mismatchdiscrimination was found to be enhanced on the surface in thecase of both LNA and DNA sensor probes, LNA being moreefficient than DNA in this respect as in solution (Table 2).

In the present scenario, as the melting reaction proceeds toachieve the equilibrium between the solution and the probeconcentrations on the gold surface, the removal of the solutionafter each heating step could cause denaturation of some moreduplexes even if at the same temperature. So the accurate Tmvalues of the surface confined duplexes would be somewhathigher than the Tm values reported herein. However, since foroligonucleotides, the difference between the nonequilibrium Tmand equilibrium Tm could be small, as reasoned by Anshelevichet al.39 and Wartell et al.,40 it is likely that the reported Tmvalues do not largely deviate from the equilibrium Tm values.Also, when a comparative view is taken, i.e., the Tm values and/or the mismatch discrimination ability of the surface-confinedLNA probes for different ionic conditions are compared, the

Figure 2. Graphs show the melting behavior of the (A) DNA-1−Cy3DNA-1 duplex (fully matched), (B) DNA-2−Cy3 DNA-1 duplex(singly mismatched) on a gold(111) surface, (C) LNA-1−Cy3 DNA-1duplex (fully matched), and (D) LNA-2−Cy3 DNA-1 duplex (singlymismatched) on a gold(111) surface.

Table 2. Differences in Melting Temperatures between FullyMatched and Singly Mismatched Situations for SolutionPhase Measurements and On-Surface Measurements

sensor probe solution (°C) surface (°C)

DNA 9.9 18.3LNA 21.5 23.6



factor of inaccuracy in the Tm values should be largely canceledout.In order to analyze the denaturation profiles of the surface-

confined LNA−DNA and DNA−DNA duplexes, the plots formelted fractions ( f) of the single stranded LNA/DNA againsttemperature (T) were obtained by fitting the melting profile ina two-state transition model that assumes that only two species(fully associated and fully dissociated) are present (Figure 3).

From the plots, it was observed that melting transition of theLNA−DNA and the DNA−DNA duplexes were quite similarand took place within a broad temperature range of about 12°C, which is likely to be due to a lack of conformationalhomogeneity of the surface-tethered LNA−DNA and DNA−DNA duplexes.The robustness of the nucleic acid layers upon storage after

first use was checked by assessing the hybridization efficiencyfor a second or third time hybridization, which was carried outon the same day and after a week’s storage using the same chip.The efficiency of both the DNA-modified chip and the LNA-modified chip reduced, though the LNA-modified chip retaineda higher degree of hybridization efficiency even after a week’sstorage (Table S1 in the Supporting Information). Theexperiments for assessing the level of hybridization efficiencyretained, after storage, were carried out for the physiologicallyrelevant salt concentration of 100 mM only.For control experiments, the thiolated LNA-3 and DNA-3

sensor strands were immobilized onto the gold(111) surfacekeeping all the sample preparation conditions the same as inthe case of the Tm measurement experiments (see theSupporting Information). The modified gold substrates werethen treated with the fully mismatched Cy3 DNAnc strands. Nofluorescence signal could be detected, indicating that nosignificant level of nonspecific adsorption of the target strandstook place in either of the two cases (see Figure S3a,b in theSupporting Information). Furthermore, in order to test whetherheating could alter the fluorescence capacity of the fluorophore-labeled oligonucleotides and thereby interfere with the Tmmeasurements, the Cy3 DNA-1 solution was heated to 70 °C,then cooled down to room temperature, and applied onto athiolated LNA-1 modified gold(111) surface. The fluorescenceimage and the intensity were found to be similar to the sampleprepared using unheated Cy3 DNA-1 solution (see FigureS3c,d in the Supporting Information). In order to check thequality of immobilization as well as to confirm whether the

reduction in fluorescence intensity with increase in temperaturewas solely due to duplex dissociation or it was influenced by thesurface probe desorption, Cy3-tagged thiolated LNA wasimmobilized onto gold substrate and heated to differenttemperatures by following the same procedure as was for allthe other experiments and fluorescence images were captured.It was observed that the produced film was more or lesshomogeneous having the probe density of about (1−10) × 1015

molecules/cm2 (from our ongoing experiments) whereas in thecase of DNA the probe density reported on the gold surfacewas about (1−2) × 1013 molecules/cm2.41 We also found thatno significant amount of thiolated immobilized probe wasdesorbed from the gold surface upon heating up to 80 °C, asthe captured fluorescence images at different temperatures werealmost identical, having very similar fluorescence intensityvalues (see Figure S4 in the Supporting Information). Earlier, inseveral studies,42,43,17,8 the thiol-DNA SAM on the gold surfacewas heated at 70 °C or beyond 70 °C for inducing adehybridization event. Peterson et al. showed that only thetarget DNA strands were removed from the gold surface whenthe surface modified with the duplex film was heated at 70/80°C.17 No thiol-DNA was lost from the surface and, moreover,the films did not loose their specificity upon heating at thattemperature.43,17 Therefore, in our study too, the reduction offluorescence emission upon heating is expected not to have anysignificant contribution from detachment of the probes fromthe surface, and it is more likely that the fluorescence reductionis primarily due to dehybridization of the surface-boundduplexes. As far as fluorescence quenching close to thegold(111) surface is concerned, the quenching effects shouldlargely remain the same for all the experiments, since similarpreparative conditions were maintained all through. If acomparative view is taken (i.e., if the Tm values at differentionic conditions or the mismatch discrimination at differentionic conditions are compared), such effects are expected to begenerally canceled out.

Tm Measurement of the LNA−DNA and DNA−DNADuplexes in Solution. The Tm profiles of all the hybridizednucleic acid samples (LNA−DNA and DNA−DNA duplexes,in both fully matched and singly mismatched combinations)were obtained by measuring absorbance over a temperaturerange of 25−96 °C. All the melting temperature profiles weremonophasic and the melting curves were sigmoidal in nature,which indicate cooperative melting behavior of the nucleic acidduplexes in solution (see Figures S5 and S6 in the SupportingInformation).

Role of Salt Concentration in “On-Surface” MismatchDiscrimination by LNA Probes. In order to assess thepossibilities of enhancement of the mismatch discriminationability of surface-tethered LNA sensor probes by ionicadjustments, we varied the concentration of NaCl in hybrid-ization buffer within the range 50−1000 mM. With an increasein salt concentration, the Tm of the fully matched duplexesincreased significantly, whereas the singly mismatched duplexeswere least affected, in the case of both LNA and DNA probes(Table 3). The interstrand repulsion between the negativelycharged strands of the fully matched duplexes could beincreasingly compensated by the positive counterions as the saltconcentration was increased. This resulted in a meltingtemperature rise. The interstrand repulsion is expected to beless severe in the case of the singly mismatched duplexes due toa relatively insignificant extent of hybridization and thereforelesser proximity of the two strands. This rendered the Tm values

Figure 3. Typical melted fraction (f) vs temperature curves for thefully matched DNA−DNA (dotted line) and LNA−DNA (solid line)duplexes on the gold(111) surface. In both the cases, immobilizationof sensor strands and subsequent hybridizations were carried out insodium phosphate buffer (20 mM sodium phosphate, 100 mM sodiumchloride, pH 7.0).



of the singly mismatched duplexes almost independent of saltconcentration, and with increase in sodium concentration from50 to 1000 mM, we observed a Tm rise of only about 1 °C. Thisobservation is consistent with that made earlier in a study ondetection of DNA sequences using DNA sensor probes,14

where it was shown that at high sodium ion concentrations(>100 mM), destabilization of the mismatched duplexes wasslightly lower in comparison to the lower sodium ionconcentrations. Effectively, such a disparity between the waysthe Tm values of the fully matched and the singly mismatchedduplexes depended on salt concentration could result inenhancement of the single base mismatch discrimination abilityof both LNA and DNA probes as the salt concentration wasincreased (Table 3). Plot of Tm vs log[Na+] was found to belinear (Figure 4A) within the salt concentration range 50−1000mM for both LNA and DNA. Positive slope of the plotsindicates counterion association during duplex formation. Theslopes were more positive for the fully matched duplexescompared to the singly mismatched duplexes indicating thatcounterion association along with duplex stabilization wassignificant for the fully matched duplexes, whereas for singlymismatched duplexes such an effect was almost negligible.Because of this reason, maximization of the single basemismatch discrimination ability of the LNA sensor layercould be achieved by varying salt concentration in thehybridization buffer. In solution phase studies on LNA−DNAduplex formation, using partially LNA-modified sequences, it

was reported that mismatch discrimination ability of LNA-modified probes was similar, whether the buffer contained 69mM Na+ or 1 M Na+.30 In the case of fully LNA-modifiedsequences, however, we found that though mismatchdiscrimination ability of LNA probes remained almost thesame for 50 and 100 mM Na+, it could be increased when theNa+ concentration was raised to 500 mM (Table S3 in theSupporting Information). On the gold(111) surface, themismatch discrimination ability of the surface-tethered LNAprobes was found to be more susceptible to ionic modulationand excelled solution behavior in this respect for the saltconcentrations ≥100 mM onward (see Table 3 and Table S3 inthe Supporting Information). We further observed that in thecase of the fully matched LNA−DNA duplexes, the effect ofchange in the salt concentration on Tm values was ratherpronounced within the salt concentration range 50−100 mM,whereas for higher salt concentration values up to 1000 mM,the Tm increased with a uniform rate (Table 3). In case of thefully matched DNA−DNA duplexes, however, the rate ofincrease in Tm values was more or less consistent throughoutthe entire range of salt concentration (50−1000 mM) applied(Table 3). This means that mismatch discrimination by LNAprobes could be relatively drastically enhanced within the saltconcentration range 50−100 mM compared to the higher saltconcentrations, while for DNA probes, mismatch discrim-ination ability was more or less uniformly affected throughoutthe entire range of salt concentration applied (Table 3). Themismatch discrimination ability of LNA probes could bemaximized for the salt concentration of 1 M (Table 3).However, with a further increase in NaCl concentration to 2and 3 M, the Tm values of the fully matched LNA−DNAduplexes reduced, the singly mismatched duplexes remainedunaffected, and therefore the mismatch discrimination ability ofthe LNA probes reduced (Figure 4B).As both LNA and DNA are negatively charged species, an

electrostatic penalty needs to be paid as the repulsiveinteractions increase when the ssLNA and the ssDNA strandscome in close proximity prior to duplex formation. This penaltycan be reduced by increasing the amount of salt and/or byadjusting the type of cation in hybridization buffer. It has beenrecently reported by Fuchs et al. that the electrostaticinteractions can clearly influence DNA−DNA duplex stabilityfor salt concentrations up to 620 mM of NaCl in the case of 16-

Table 3. Melting Temperatures of Nucleic Acid Duplexes onGold(111) Surface for Different Sodium ChlorideConcentrations in 20 mM Sodium Phosphate Buffer (pH7.0)a

NaCl concn (mM)

nucleic acid duplexes 50 100 500 1000

DNA-1−Cy3 DNA-1 44.7 °C 47.8 °C 53.9 °C 57.8 °CDNA-2−Cy3 DNA-1 29.3 °C 29.5 °C 30.0 °C 30.2 °CΔTm (DNA) 15.4 °C 18.3 °C 23.9 °C 27.6 °CLNA-1−Cy3 DNA-1 41.2 °C 52.7 °C 57.4 °C 61.1 °CLNA-2−Cy3 DNA-1 28.4 °C 29.1 °C 29.2 °C 29.4 °CΔTm (LNA) 12.8 °C 23.6 °C 28.2 °C 31.7 °C

aDifferences in melting temperatures between fully matched and singlymismatched situations are shown as ΔTm.

Figure 4. Variation of melting temperature (Tm) of the LNA−DNA duplexes with (A) log [Na+] and (B) with [NaCl] on the gold(111) surfacecompared to corresponding DNA−DNA duplexes. In part A, the empty symbols correspond to DNA−DNA duplexes and the filled symbolscorrespond to LNA−DNA duplexes, while in both of the cases the circles symbolize fully matched and the squares symbolize a single base mismatchsituation. In part B, the solid circles symbolize fully matched LNA−DNA duplexes, and the solid squares symbolize singly mismatched LNA−DNAduplexes.



mer thiol-DNA strands self-assembled on the gold(111)surface.44 In the present study, we varied the NaClconcentration in the hybridization buffer from 50 mM to 3M. It was observed that the Tm of the duplexes increased forsalt concentrations up to 1 M. This fact can be explained by theshielding effect of the sodium ion that can minimize theelectrostatic repulsion between the negatively charged LNA andDNA backbones, thereby stabilizing the duplexes. Beyond 1 Mhowever, the Tm was found to reduce. It has been reportedearlier that at high salt concentration (NaCl > 1 M) where theelectrostatic contributions saturate, the presence of backbonecharges can no longer influence the interactions.45 However,since there can be water structure rearrangement at the water−solute interface as a result of high salt concentration,46,47 saltcan still influence the stability of the macromolecular structureindirectly at high salt concentrations. In the present case, thedecrease in melting temperature for salt concentrations beyond1 M could most likely be a result of destabilization of theduplexes due to water structure rearrangement.Role of Cationic Charge in “On-Surface” Mismatch

Discrimination by LNA Probes. In order to investigate howthe charge of cation could affect the melting behavior of theLNA−DNA duplexes, we incorporated magnesium ion (2+),spermidine (3+), and spermine (4+) in the hybridization bufferat varied concentrations. It was observed that with an increasein the magnesium ion (Mg2+) concentration, the Tm values ofthe fully matched LNA−DNA duplexes increased noticeably,whereas for the singly mismatched duplexes, the increase in Tmwas negligible (Table 4), effectively resulting in enhancedmismatch discrimination with increasing Mg2+ concentration.Similar trends, i.e., clear increase in Tm of the fully matchedduplexes and almost no change in Tm of the singly mismatchedduplexes, were observed in the case of Na+ treatment too(Table 3). Within the concentration range of 10−15 mM, theeffect of Mg2+ on Tm of the fully matched LNA−DNA duplexeswas most pronounced. Below 10 mM, the increase in Tm withincreasing salt concentration was much less prominent (datanot shown). Above 15 mM, the effect of salt concentrationseemed to become saturated. It is evident from the Tm valuesthat a significantly lower concentration of the divalentmagnesiun ion could be applied compared to the concentrationof monovalent sodium ion for achieving a similar level of thestability of the LNA−DNA duplexes (Tables 3 and 4). It hasbeen shown earlier that the Mg2+ distribution around anisolated DNA duplex is more compact than the Na+

distribution.48 Therefore, the total sum of positive charges ishigher for magnesium under the restricted spatial conditions,leading to a stronger stabilization effect of Mg2+ over Na+.14 Inthe case of LNA, a similar situation might be prevailing since ithas been shown that the mode of binding of Mg2+ to nucleicacids depends essentially on the backbone negative chargedensity.49 As LNA residues do not introduce any significantcharges, this result is quite expected. The single base mismatch

discrimination ability of the LNA sensor film could bemaximized at a 15−20 mM Mg2+ concentration. In comparison,the effect of Na+ was most pronounced for the concentrations50−100 mM. Maximization of mismatch discrimination couldtherefore be better achieved by applying Mg2+. In the solutionphase study, it was reported that the interactions of Na+ andMg2+ ions with probe−target duplexes are not significantlymismatch-specific.30 However, on the surface, we observed thatthe singly mismatched duplexes remained almost unaffectedcompared to the fully matched duplexes whatever were the saltconcentration and the type of cation. This indirectly impliesthat on the surface the interactions of Na+ and Mg2+ with theprobe−target duplex is mismatch-sensitive and that might arisefrom the spatial restrictions due to higher probe density on thesurface. In the presence of spermidine and spermine in thehybridization buffer, we found almost the same results as weobtained for Mg2+ except for the fact that the results wereobtained with lower concentrations of the polyaminescompared to Mg2+. However, in the case of DNA, it wasshown by Hou et al. that polyamines, especially spermine, canstabilize the mismatched duplex to such an extent that it couldbe as stable as the perfectly matched duplex.50 So it is quiteobvious that the mode of interaction of spermine with thesurface-tethered LNA−DNA duplexes differ in some aspectscompared to the DNA−DNA duplexes. That might arise fromthe lesser backbone flexibility of the LNA molecules as well asspatial restrictions on the surface.It was observed that there is minimal hysteresis present

between the solution heating and the cooling curves (Figure S6in the Supporting Information), which indicates the largelyreversible character of the melting transitions. Given thiscondition of reversibility to have met in our experiments, it islikely that the observed differences in melting temperaturebetween the fully matched and the mismatched duplexes andbetween LNAs and DNAs are primarily determined by thefactors related to base stacking and H-bonding. Enhanced basestacking and H-bonding in the case of the fully matchedduplexes compared to the singly mismatched duplexes (wherethe mismatch is centrally placed) should primarily control thedifferences in melting temperature between fully matched andsingly mismatched duplexes. In order to account for thedifferences between LNAs and DNAs, it has been observedearlier in solution phase experiments that single base mismatchdiscrimination can be performed better using the LNA probes,compared to the DNA probes.24,2 This superior discriminationcapability of LNA has been attributed to the fact that LNAmodification enhances base stacking of fully matched base pairsand decreases stabilizing stacking interactions of mismatchedbase pairs.30 This trend that is observed in the solution phaseexperiments has been reflected in the case of the immobilizedLNA strands onto the gold(111) surface as shown by Mishra etal.23 The trend could be reproduced on the surface probablybecause of the structural rigidity of the LNA backbone that

Table 4. Melting Temperatures of LNA−DNA Duplexes on Gold(111) Surface for Different Magnesium Chloride, Spermidine,and Spermine Concentrations in 20 mM Sodium Phosphate Buffer (pH 7.0)a

MgCl2 concn (mM) spermidine concn (mM) spermine concn (mM)

nucleic acid duplexes 10 15 20 5 10 15 5 10

LNA-1−Cy3 DNA-1 48.5 °C 62.7 °C 62.9 °C 53.9 °C 63.2 °C 54.5 °C 56.8 °C 63.8 °CLNA-2−Cy3 DNA-1 29.9 °C 30.1 °C 30.6 °C 34.5 °C 33.6 °C 33.5 °C 34.1 °C 34.3 °CΔTm 18.6 °C 32.6 °C 32.3 °C 19.4 °C 29.6 °C 21.0 °C 22.7 °C 29.5 °C

aDifferences in melting temperatures between fully matched and singly mismatched situations are shown as ΔTm.



could result in reduced nonspecific interactions with thesurface28 helping the LNA strand to adopt a more uprightorientation and therefore making it least susceptible to thesurface effects. The upright orientation also makes the LNAstrand more accessible to the target DNA strand, ensuringgreater fidelity in sequence recognition and therefore superiorsingle base mismatch discrimination. Piao et al. reported thatLNA purines offer great potential to recognize mismatches thanLNA pyrimidines and DNA purines.38 In addition to that,Owczarzy et al. reported that a centrally placed A.G mismatch,in the case of LNA, has the largest discriminatory boost.37 Inour case, since there is a centrally placed A.G mismatch (Table1), enhanced mismatch discrimination in the case of the LNAprobes, compared to the DNA probes, is expected forthermodynamic reasons.Melting of a nucleic acid duplex in solution means total

unzipping of the duplex, where the individual strands can freelymove away from each other. The same pathway is not expectedto be followed within a film at a solid−liquid interface sinceeach duplex is closely surrounded by a number of otherduplexes within a film structure. The local concentration of thenegatively charged oligonucleotides and therefore the probedensity could be considerably higher on the surface comparedto that in bulk solution since the probes can be chemicallyanchored on the surface. This allows both the steric effects aswell as the electrostatic effects to play an important role insolid-phase hybridization. Interpretation of experimental dataon hybridization/dehybridization events occurring in a film atthe solid−liquid interface can pose a considerable amount ofchallenges and is often not straightforward. A number ofsources of deviation from ideal behavior need to be considered.For example, immobilization may not be best achieved for allthe sensor probes due to the presence of surface defects;second, the sensor probes may not all be oriented suitably fortarget-binding due to inhomogeneities in the film structureand/or nonspecific interactions with the surface, e.g., gold−nitrogen (nucleobase) interactions; third, the probe densitymay be too high that would inhibit effective target entry intothe sensor film prior to hybridization. As a consequence of acomplex interplay of such deviations from a “solution-like”situation, the “on-surface” Tm values are unlikely to follow thesame pattern of change as in the case of solution measurements.The thermal melting experiments on surface-confined nucleicacid duplexes can therefore provide only an approximate ideaabout the thermodynamic stability of the nucleic acid duplexes.Despite the possible sources of ambiguities as described, twoclear conclusions could be made in the present study. First,single base mismatch discrimination could be better performedby LNA probes compared to DNA probes on the gold(111)surface. Second, the single base mismatch discrimination abilityof the LNA probes could be enhanced by increasing themonovalent salt (NaCl) concentration to 1 M. Mismatchdiscrimination could also be improved by using polyvalentcations like magnesium ion (2+), spermidine (3+), andspermine (4+), the maximum by a magnesium ion consideringcomparable concentrations in each case (Table 4). Importantly,we found that the same extent of duplex stabilization could beachieved by applying nearly 2 orders of magnitude lowerconcentrations of the polyvalent cations compared to theconcentration of monovalent sodium ion. Unlike DNA, LNAhas the modified ribose moiety which reduces the conforma-tional flexibility and increases local organization of thephosphate backbone,51 and the ssLNA strands could avoid

nonspecific interactions with gold surface and be oriented moreupright with respect to the surface.23 It is likely that thisorientational advantage of the ssLNA strands could positivelyinfluence its mismatch discrimination capacity compared to thessDNA strands.

■ CONCLUSIONSIn conclusion, we report for the first time how the single basemismatch discrimination ability of surface-anchored LNAprobes can be maximized by ionic adjustments. Whilemismatch discrimination by both DNA and LNA probes wasfound to be improved on the surface, in comparison to thesolution phase, LNA probes excelled DNA probes both insolution and on the surface. Both the type and theconcentration of cations were found to play a decisive role insolid phase LNA−DNA hybridization. As the effect of saltconcentrations on singly mismatched duplexes was almostnegligible compared to the fully matched duplexes, single basemismatch discrimination ability of the LNA sensor layer couldbe enhanced with an increase in salt concentration. When wecompare these results to the behavior of DNA sensor probes, itis revealed that ionic tuning results in more prominent effectson LNA-based single base mismatch discrimination comparedto the DNA-based discrimination. This means that LNA probesare more susceptible to changes in the ionic environment andcan therefore respond to smaller changes in ionic conditions.

■ ASSOCIATED CONTENT*S Supporting InformationAdditional information as noted in text. This material isavailable free of charge via the Internet at http://pubs.acs.org.

■ AUTHOR INFORMATIONCorresponding Author*Phone: +91 33 2473 4971 ext. 1506. Fax: +91 33 2473 2805.E-mail: [email protected] authors declare no competing financial interest.

■ ACKNOWLEDGMENTSWe acknowledge financial support (Grant No. BT/PR-11765/MED/32/107/2009) from DBT, Government of India, andresearch fellowships of S.M. and S.G. from CSIR, Governmentof India.

■ REFERENCES(1) Barnes, M. R. Human Genetic Variation: Databases andConcepts. In Bioinformatics for Geneticists; Barnes, M. R., Gray, I. C.,Eds.; John Wiley and Sons, Ltd.: Chichester, U.K., 2003; Chapter 3.(2) Ugozzoli, L. A.; Latorra, D.; Arar, K.; Hamby, K. Anal. Biochem.2004, 324, 143−152.(3) Johnson, M. P.; Haupt, L. M.; Griffiths, L. R. Nucleic Acids Res.2004, 32, e55.(4) Wang, L.; Blasic, J. R.; Holden, M. J., Jr.; Pires, R. Anal. Biochem.2005, 344, 257−265.(5) Chou, L.-S.; Meadows, C.; Wittwer, C. T.; Lyon, E.BioTechniques. 2005, 39, 644−647.(6) Mouritzen, P.; Nielsen, A. T.; Kongsbak, L.; Møller, S. Expert Rev.Mol. Diagn. 2003, 3, 27−38.(7) Kauppinen, S.; Nielsen, P. S.; Møller, S.; Ramsing, N. B.PharmaGenomics 2003, 3, 24−34.(8) Du, H.; Strohsahl, C. M.; Miller, B. L.; Krauss, T. D. J. Am. Chem.Soc. 2005, 127, 7932−7940.



http://pubs.acs.org

mailto:[email protected]

(9) Sassolas, A.; Leca-Bouvier, B. D.; Blum, L. J. Chem. Rev. 2008,108, 109−139.(10) Levicky, R.; Horgan, A. Trends Biotechnol. 2005, 23, 143−149.(11) Mir, K. U.; Southern, E. M. Annu. Rev. Genomics Hum. Genet.2000, 1, 329−360.(12) Homola, J. Chem. Rev. 2008, 108, 462−493.(13) Zhang, L.; Wu, C. L.; Carta, R.; Zhao, H. T. Nucleic Acids Res.2007, 35, e18.(14) Springer, T.; Sípova, H.; Stepanek, J.; Homola, J. Nucleic AcidsRes. 2010, 38, 7343−7351.(15) Yu, F.; Yao, D. F.; Knoll, W. Nucleic Acids Res. 2004, 32, e75.(16) Okahata, Y.; Kawase, M.; Furusawa, H.; Ebara, Y. Anal. Chem.1998, 70, 1288−1296.(17) Peterson, A. W.; Heaton, R. J.; Georgiadis, R. M. Nucleic AcidsRes. 2001, 29, 5163−5168.(18) Cho, Y. K.; Kim, S.; Pak, Y. E.; Ha, T. H.; Kim, K. J. ColloidInterface Sci. 2004, 278, 44−52.(19) Fritz, J.; Baller, M. K.; Lang, H. P; Guntherodt, H.-J.; Gerber,Ch.; Gimzewski, J. K. Science 2000, 288, 316.(20) Mukhopadhyay, R.; Lorentzen, M.; Kjems, J.; Besenbacher, F.Langmuir 2005, 21, 8400.(21) Drummond, T. G.; Hill, M. G.; Barton, J. K. Nat. Biotechnol.2003, 21, 1192.(22) Hansen, J. A.; Mukhopadhyay, R.; Hansen, J. O.; Gothelf, K. V.J. Am. Chem. Soc. 2006, 128, 3860.(23) Mishra, S.; Ghosh, S.; Mukhopadhyay, R. Langmuir 2012, 28,4325−4333.(24) Singh, S. K.; Nielsen, P.; Koshkin, A. A.; Wengel, J. Chem.Commun. 1998, 455.(25) Koshkin, A. A.; Nielsen, P.; Singh, S. K.; Wengel, J. J. Am. Chem.Soc. 1998, 120, 13252.(26) Vester, B.; Wengel, J. Biochemistry 2004, 43, 13233.(27) Kurreck, J.; Wyszko, E.; Gillen, C.; Erdmann, V. A. Nucleic AcidsRes. 2002, 30, 1911.(28) Martinez, K.; Estevez, M.-C; Medley, C. D.; Tan, W. Anal.Chem. 2009, 81, 3448.(29) Pande, V.; Nilsson, L. Nucleic Acids Res. 2008, 36, 1508.(30) You, Y.; Moreira, B. G.; Behlke, M. A.; Owczarzy, R. NucleicAcids Res. 2006, 34, e60.(31) Proll, F.; Mohrle, B.; Kumpf, M.; Gauglitz, G. Anal. Bioanal.Chem. 2005, 382, 1889−1894.(32) Owczarzy, R. Biophys. Chem. 2005, 117, 207−215.(33) Ghosh, S.; Mukhopadhyay, R. J. Colloid Interface Sci. 2011, 360,52.(34) Hianik, T.; Gajdos, V.; Volkov, E.; Vadgama, P. Bioelectrochem-istry 2001, 53, 199−204.(35) Lee, J. W.; Sima, S. J.; Choa, S. M.; Lee, J. Biosens. Bioelectron.2005, 20, 1422−1427.(36) Nuzzo, R. G.; Allara, D. L. J. Am. Chem. Soc. 1983, 105, 4481−4483.(37) Owczarzy, R.; You, Y.; Groth, C. L.; Tataurov, A. V. Biochemistry2011, 50, 9352−9367.(38) Piao, X. Y.; Yan, Y.; Yan, J.; Guan, Y. Anal. Bioanal. Chem. 2009,394, 1637−1643.(39) Anshelevich, V. V.; Vologodskii, A. V.; Lukashin, A. V.; Frank-Kamenetskii, M. D. Biopolymers 1984, 23, 39−58.(40) Wartell, R. M.; Benight, A. S. Phys. Rep. 1985, 126, 67−207.(41) Demers, L. M.; Mirkin, C. A.; Mucic, R. C.; Reynolds, R. A.;Letsinger, R. L.; Elghanian, R.; Viswanadham, G. Anal. Chem. 2000, 72,5535−5541.(42) Peterlinz, K. A.; Georgiadis, R. M. J. Am. Chem. Soc. 1997, 119,3401−3402.(43) Herne, T. M.; Tarlov, M. J. J. Am. Chem. Soc. 1997, 119, 8916−8920.(44) Fuchs, J.; Fiche, J.−B.; Calemczuk, R.; Livache, T. Biophys. J.2010, 99, 1886−1895.(45) Arakawa, T.; Timasheff, S. N. Biochemistry 1984, 23, 5912−5923.

(46) Hamaguchi, K.; Geiduschek, E. P. J. Am. Chem. Soc. 1962, 84,1329−1338.(47) Collins, K. D.; Washabaugh, M. D. Q. Rev. Biophys. 1985, 18,323−422.(48) Misra, V. K.; Draper, D. E. J. Mol. Biol. 1999, 294, 1135−1147.(49) Anderson, C. F.; Record, M. T., Jr. Annu. Rev. Phys. Chem. 1995,46, 657−700.(50) Hou, M.; Lin, S.; Wang, A. H. J.; Kan, L. Nucleic Acids Res. 2001,29, 5121−5128.(51) Braasch, D. A.; Corey, D. R. Chem. Biol. 2001, 8, 1−8.



Documents

Maximizing Mismatch Discrimination by Surface-Tethered Locked Nucleic Acid Probes via Ionic Tuning