Research Article Microarrays as Model Biosensor Platforms ...Research Article Microarrays as Model Biosensor Platforms to Investigate the Structure and Affinity of Aptamers JenniferA.Martin,

Research ArticleMicroarrays as Model Biosensor Platforms toInvestigate the Structure and Affinity of Aptamers

Jennifer A Martin12 Yaroslav Chushak13 Jorge L Chaacutevez12

Joshua A Hagen1 and Nancy Kelley-Loughnane1

1Human Effectiveness Directorate 711 Human Performance Wing Air Force Research LaboratoryWright-Patterson Air Force Base OH 45433 USA2UES Inc 4401 Dayton-Xenia Road Dayton OH 45432 USA3The Henry M Jackson Foundation for the Advancement of Military Medicine 6720A Rockledge Drive Bethesda MD 20817 USA

Correspondence should be addressed to Nancy Kelley-Loughnane nancykelley-loughnane1usafmil

Received 12 November 2015 Accepted 17 January 2016

Academic Editor Satoshi Obika

Copyright copy 2016 Jennifer A Martin et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Immobilization of nucleic acid aptamer recognition elements selected free in solution onto the surface of biosensor platforms hasproven challenging This study investigated the binding of multiple aptamertarget pairs immobilized on a commercially availablemicroarray as a model system mimicking biosensor applications The results indicate a minimum distance (linker length) fromthe surface and thymine nucleobase linker provides reproducible binding across varying conditions An indirect labeling methodwhere the target was labeled with a biotin followed by a brief Cy3-streptavidin incubation provided a higher signal-to-noise ratioand over two orders of magnitude improvement in limit of detection compared to direct Cy3-protein labeling We also showedthat the affinities of the aptamertarget interaction can change between direct and indirect labeling and conditions to optimizefor the highest fluorescence intensity will increase the sensitivity of the assay but will not change the overall affinity Additionallysome sequences which did not initially bind demonstrated binding when conditions were optimizedThese results in combinationwith studies demonstrating enhanced binding in nonselection buffers provided insights into the structure and affinity of aptamerscritical for biosensor applications and allowed for generalizations in starting conditions for researchers wishing to investigateaptamers on a microarray surface

1 Introduction

Aptamers are oligonucleotide sequences that bind to aspecific target molecule through noncovalent interactions[1 2] Aptamers have been implemented in a wide varietyof applications including analytical purification reagentshistological detection reagents targeted delivery vehiclesand pharmaceuticals [3ndash6]Moreover aptamers have demon-strated particular promise as surface-immobilized biosensorrecognition elements due to their stability their ease ofchemical modification and high achievable densities whentethered to a platform One critical issue associated withinterfacing an aptamer with biosensor detection technologystems from the typical methodology used to initially identifythe recognition elements In SELEX (Systematic Evolution

of Ligands by EXponential Enrichment) aptamers are gen-erated by iterations of incubating oligonucleotides in freesolution with targets followed by amplification over multiplerounds of selection [7 8] The solution-selected aptamer isthen synthesized with a chemical tag for immobilization ona solid support for biosensor development Previous studieshave shown that the affinity of the aptamer for the target maybe significantly diminished or destroyed after surface immo-bilization potentially due to interactions with the surfaceand other probes denaturing or occluding the binding sitesteric constraints that prevent proper aptamer folding andortarget interaction or electrostatic repulsion of the target bythe surface [9ndash11]Therefore understanding the performanceand underlying structural features of an aptamer tethered toa surface is an important aspect influencing sensor design

Hindawi Publishing CorporationJournal of Nucleic AcidsVolume 2016 Article ID 9718612 11 pageshttpdxdoiorg10115520169718612

2 Journal of Nucleic Acids

Microarray technology can be used to probe the func-tional characteristics of aptamers selected by solution-basedSELEX in an immobilized format mimicking the ultimatebiosensing application Microarrays indicate target bindingby identifying locations of fluorescence from a fluorescentlylabeled target and correlating them with the position ofknown sequences on the array Early attempts in microarraydesign suffered from challenges related to oligonucleotidedeposition by spotting presynthesized probes onto an acti-vated array surface specifically in reproducibility whenconditions (humidity buffer etc) varied from batch to batch[12 13] Oligonucleotide spotting also rendered testing ofthousands of sequences cost prohibitive due to the expenseof synthesizing each probe individually prior to depositiononto the array Current microarray technology based onsynthesis from the array surface rather than oligonucleotidespotting produces high reproducibility and allows for up to106 completely custom 80 mer DNA sequences synthesizedper array The massively parallel nature of microarray testingdrastically reduces the time and cost investments associatedwith assaying the functional characteristics of aptamerscompared to traditionalmethods of individually synthesizingand purifying each sequence of interest For example amicroarray experiment can be completed in less than oneday with data analysis requiring another 1-2 days A singlemillion feature array (Agilent 1times1M) similar to those usedin the current study at a price of sim$600 can assay up toa million sequences simultaneously compared to sim$135 forone sequence HPLC purified from commercial vendors fortraditional analysis

Onemajor challenge associated with applyingmicroarraytechnology to aptamer studies lies in the fact that current pro-tocols have been mostly developed for gene expression stud-ies that function by direct hybridization of the fluorescent-tagged genomic fragments to cDNA on the microarray Thistype of interaction forms an extremely stable duplex betweenthe base pairs of the two strands whereas aptamerproteintarget interactions may heavily rely on one or a fewnoncovalent interactions For example the binding of theimmunoglobulin E (IgE) aptamer has been shown to be com-pletely destroyed following change of a single base within theaptamer sequence [11] Moreover the majority of dye systemsused to label protein targets (ldquodirect labelingrdquo) formicroarraywork possess conjugated 120587-systems reported to interact withDNA directly [14] This observation renders it challengingto differentiate between sequences which are binding theactual protein target and those that instead interact with thedye Gold and coworkers reported using photoactive basesubstitutions to covalently cross-link proteins to the arraysfollowed by stringent washing steps to remove nonspecificbinders [15] However commercial microarray manufactur-ers do not currently offer these base substitutions in theirarray synthesis portfolio A feasible method of biotinylatingthe protein followed by a short dye-labeled streptavidinincubation (indirect labeling) demonstrated less nonspecificbinding compared to direct dye labeling [14 16] Further-more many parameters will affect the aptamer response thatmay not be as significant in cDNA hybridizations including

probe density proximity of the binding site to the surfaceimmobilization orientation (51015840 or 31015840 end) and the bufferused in the studies [10 17]Therefore protocols and methodsfor aptamers must be developed and optimized by theindividual researcher accounting for challenges specific to theapplication

As is the case with biosensor applications the responseof an aptamer in solution can be significantly different thanwhen it is attached to a microarray surface [9] Severalstudies report differences in aptamer affinity when com-paring microarray and solution-based dissociation constantmeasurements [10 11] In most cases the affinities of theseaptamers were significantly lower on the microarray thanwhen tested in the solution-based conditions used duringtheir selection [18] Significant work has been performed bydifferent groups in the use of aptamers for target detection incustom-mademicroarray settings It has been confirmed thatknown DNA aptamers can be immobilized on a microarrayand conditions to preserve their activity can be found andthe conditions used for these studies should mimic theconditions used during the initial solution-based selectionprocess [12 18] Prior work has also demonstrated thatextending the distance of the aptamers from the arraysurface enhances target binding fluorescence intensity isproportional to the concentration of target added and higherrelative fluorescence values between probes on the samearray indicate higher affinity interactions [11 16] Substan-tial work has also been presented describing the ability ofmicroarrays to assay target binding after aptamer mutationstruncations or changing conditions but these studies haveeither not analyzed multiple aptamers binding multipletargets or not utilized specific conditions to analyze struc-tural and functional information regarding target binding[11 18 19]

The main goal of the current studies was to investigatemicroarrays as a model system to examine aptamer bindingin an immobilized format mimicking that of biosensorapplications During the course of these studies several dif-ferent conditions were investigated related to the functionalcharacteristics of multiple aptamers on custom microarraysThese enabled generalizations related to potential startingpoints for researchers wishing to utilize microarrays in futureaptamer studies as well as providing insights into specificaptamertarget interactions Specifically we investigated theeffect of linker length with IgE as a control and expanded thework to analyze the effect of the linker identity dissociationconstant (119870

119889) under varying conditions effect of the dye-to-

protein ratio (DP) and comparison of a direct and indirectlabeling method [11] The schematic in Figure 1 depicts asummary of the variables investigated using microarrays inthe current study In addition the linker length and identityeffects were extended to include several guanine-quartet (G-quartet) aptamers in relation to thrombin and dye bindingFurthermore we utilized the known ability of G-quartets tobind 120587-conjugated dye systems to assess the structural char-acteristics of G-quartet aptamers under different conditionsThis enables a correlation between the conditions and theaptamer structure without concern regarding whether theselected conditions may impart conformational changes on

Journal of Nucleic Acids 3

Direct labeling method Indirect labeling method

Fluorescent-labeled target

extraction

Biotinylatedtarget

Dye-labeled streptavidin

(short)

extractionLocation Sequence FluorescenceA1 28A2 6A3 16456A4 137A5 5361A6 10679

TTT

TTT

TTTT

TTTTT

TTTTTT

TTTT

AAAA

CCCC

GGGG

Variable linker length

Variable linker identity

Adenine (A)Cytosine (C) Thymine (T) Data

Data

Guanine (G)

N

N

NN

HN

H

HNN

NN

N

N

H

H

NH2

NH2

NH

OO

O

O

H2N

AATGACTCG

CTAGAGTCA

ACGGGTACT GGTACGATA

CAGAAAATC

TTTACTATAG

Figure 1 Diagram of variables investigated in the current experiment Linker length the distance from the microarray surface wasvaried linker identity nucleobase (adenine cytosine guanine or thymine) was used to tether aptamers to the surface direct labelingmethod fluorescent reporter dye conjugated directly to the target molecule is incubated with aptamer microarray indirect labeling methodbiotinylated-target is added to microarray followed by a short dye-streptavidin incubation with aptamer microarray Following incubationall microarrays were washed and imaged and the data was extracted in order to link fluorescence intensity location with aptamer sequenceNucleobase structures were created using VIDA part of OpenEye software

a protein target Novel information on the structural andbinding properties of the various aptamers used in this studywas discovered as well as experimental evidence backingthe previously proposed structures of well reported thrombinand riboflavin (RB) aptamers These studies collectivelyprovide an indication that a minimal linker length may benecessary for reliable binding data and manipulation of theconditions can promote the interaction of weak bindingsequences with their target where bindingwas not previouslyobserved We propose that microarrays are an advantageousmeans to rapidly screen aptamer function under variousconditions of interest including different surface linkagesand buffer and salt conditions The results described hereindicate that microarrays can be used to probe the structuralfolding of aptamers specifically whether they form a G-quartet by a simple method of introducing a compatible dyeto the system This work is beneficial in the biosensor fieldwhere transitioning a solution-selected aptamer to a surface-immobilized sensing platform has proven to be a nontrivialtask

2 Materials and Methods

21 Chemicals and Equipment IgE (Fitzgerald) IllustraNAP-25 desalting columns and Cy3 Mono-Reactive DyePack (GE Healthcare) NanoDrop (Thermo Scientific) andnuclease freewater (Gibco)were usedMicroarray equipmentis as follows custom 8times15k and 1times1M DNA microarrays8times15k and 1times1M gasket slides ozone barrier slides hybridiza-tion chambers scanner cassettes hybridization oven andHigh-resolution Microarray Scanner (all Agilent) Slide rackand wash dishes (Shandon) and Kimtech polypropylenewipes (Kimberly-Clark) FluoReporter Mini-Biotin-XX Pro-tein Labeling Kit (Invitrogen) Streptavidin-Cy3 Conjugateand HABAAvidin Reagent (Sigma) and dialysis cassettes(35k MWCOThermo Scientific) were also used

22 Buffers Buffers are as followsbinding [PBSMTB] 1x PBS (81mM Na

2HPO4

11 mM KH2PO4 27mM KCl 137mM NaCl pH 74)

+ 1mMMgCl2+ 01 Tween-20 and 1 BSA


washing [PBSM] 1x PBS (81mM Na2HPO4 11 mM

KH2PO4 27mMKCl 137mMNaCl pH 74) + 1mM

MgCl2

rinse [R] 14 dilution of PBSM and nuclease freewater

IgE binding buffer without K+ (81mM Na2HPO4

137mM NaCl pH 74) + 1mMMgCl2+ 01 Tween-

20 and 1 BSA

washing 81mM Na2HPO4 137mM NaCl pH 74 +

1mMMgCl2

RBWash Buffer 20mMHEPES 300mM KCl 5mMMgCl

2

RB binding buffer 20mM HEPES 300mM KCl5mMMgCl

2+ 01 Tween-20 + 1 BSA

23 Cy3-Labeling of IgE and Thrombin IgE was diluted to1mgmL with 01M sodium carbonate buffer (pH = 93)One mL of 1mgmL IgE was added to one Cy3 dye packand incubated 30min Free dye was separated from IgE-bound dye by purification with a desalting column Theproduct was then injected into dialysis cassettes for twoseparate 2 h incubations with fresh 400mL PBSMTB at roomtemperature and then overnight with 400mL fresh PBSMTBat 4∘C Dye-to-protein ratio (DP) was calculated usingthe manufacturersrsquo instructions with the aid of NanoDropUVVis detection

Thrombin was labeled with Cy3 in a similar manner toIgEThrombin was incubated for 30 minutes after dilution in1mL01MNa

2CO3buffer (pH=93) to 1mgmLTheproduct

was purified with a Texas Red protein labeling size exclusioncolumn (Molecular Probes) to yield a DP = 08 using themanufacturerrsquos instructions viaNanoDropUVVis detection

24 8times15k Microarray Design and Methodology Aptamerswere synthesized on custom 8times15k microarray slides in repli-cates ranging from 5 to 15 by Agilent Technologies Blockingwith PBSMTBwas performed on theDNAmicroarray loadedinto a 50mL conical tube for 1 h at room temperature Slideswere disassembled in PBSM buffer and rinsed for 5min atroom temperature using a stir plate The slides were quicklyedge-tapped to remove excess buffer Seventy 120583L protein(variable concentrations) in PBSMTB was loaded onto thegasket slide then incubated with each of the 8 DNA arrays for2 hrs at 20∘C in a hybridization chamberhybridization ovenSlides were disassembled in PBSMTB buffer and washed for3min in a separate PBSM buffer with the slide rack andstir plate then the slide rack was transferred to 14 PBSMbufferwater for 1min using a stir plate Slides in the slide rackwere dipped in nuclease free water to remove any remainingsalt and washed for 1min with stir plate The rack was slowlywithdrawn from the water to promote a drier surface theback of the slide was wiped with ethanol and then placed in a50mL conical tube with a polypropylene wipe at the bottomand centrifuged at 4150 rpm for 3min The microarray wasloaded into a scanner cassette and covered with an ozonebarrier slide before scanning

25 1times1M Microarray The basic protocol was the same forthe 1times1M arrays as for the 8times15k arrays described above Theonly difference in methodology was that the gasket slide helda 750120583L volume instead of 70120583L

26 Indirect Labeling Method IgE Biotinylation Biotinylationof IgE was carried out according to the manufacturerrsquosinstructions and purified using the provided spin columnBriefly 200120583L 1mgmL IgE was mixed with 20120583L 1MNaHCO

3 pH = 832The biotin-XXmaterial was dissolved in

200120583L nuclease free water and 2 120583L solution was transferredto the proteinNaHCO

3mixtureThe solution was stirred for

1 hour at room temperature and then added to the preparedspin columnThe biotinprotein ratio was calculated to be 06according to the HABAAvidin test

27 8times15k and 1times1M Microarrays (Indirect Labeling Method)The 8times15k and 1times1M microarrays were analyzed in the samemanner as the direct labeling method with some exceptionsFollowing incubation with biotin-IgE slides were disassem-bled in PBSMTB buffer and rinsed for 3min on a shakerSlides were transferred to PBSM buffer for 3min also on ashaker Then 05 10 or 500 nM Cy3-streptavidin (Cy3-SA)was loaded into the gaskets and incubated for 2min Thewashing steps were the same as the direct labeling method

28 8times15k Arrays with Variable Targets The basic protocolwas similar to the 8times15k direct labeling protocol exceptthat the samples and concentrations were variable 100 nMsamples Cy3-IgE DP = 24 041 and 028 Cy5 Cy3-thrombin 50 nM Cy3 10 nM Cy3-streptavidin When thebuffers were varied the arrays were incubated in the buffer ofinterest before sample loading and slides were disassembledin water then quickly transferred (3 seconds) to PBSM for thesubsequent washing steps

29 Data Analysis The arrays were scanned using AgilentScan Control software generating TIFF images using 20-bitimaging at 3120583m (for 1times1M arrays) or 5120583m (8times15k arrays)Data was extracted using Agilent Feature Extraction softwareversion 10731 Code was written in Perl to provide meanfluorescence intensity and standard deviation of the replicatesfor each probe investigated

3 Results and Discussion

31 Linker Length Testing Using IgE The current study testedthe T linker response for the initial IgE binding aptamerIgE 17-4 at 100 nMCy3-IgE and observed an increase in flu-orescence response from linker length 0ndash15 T (Figure 2 [20])This agrees with prior studies demonstrating that increas-ing the (thymine) T-linker length separating the aptamerfrom the surface of the microarray generated an increasein sensor response [11] The initial IgE sequence IgE 17-4was mutated by Katilius et al on a microarray in a seriesof single or double mutations to generate new sequenceswhich retained target binding capabilities [11] The currentstudy showed that positive mutants IgE D17-4 DM1 37 and

Journal of Nucleic Acids 5Fl

uore

scen

ce in

tens

ity

IgE_D17-4IgE_D17-4_DM1_37IgE_D17-4_SM1

ID17-4DM13_25

ID17-4SM17

ID17-4SM21

ID17-4SM28

SA

50 15 2010

Linker length

0

100

200

300

400

500

Figure 2 Response of different T linker sequences to 100 nM Cy3-IgE in PBSMTBbuffer Error bars represent SEMof raw fluorescenceintensity values for 5ndash15 replicates of each sequence

IgE D17-4 SM1 demonstrated similar trends but displayedlower fluorescence intensity values at shorter linker lengthsthan IgE 17-4 (See Table S1 for a description of sequencesused in the current study) (see SupplementaryMaterial avail-able online at httpdxdoiorg10115520169718612) As thelinker length was extended the two mutants produced fluo-rescence intensity values approaching that of the initial IgEaptamer (Figure 2) Also several mutations that were shownin theKatiliuswork to reduce binding (ID17-4DM13 25 ID17-4SM17 ID17-4SM21 and ID17-4SM28) demonstrated fluores-cence intensities similar to the initial aptamer andhigher thanthe two binding mutants (SM1 and DM1 37) at short linkerlengths [11]However the reduced bindingmutants decreasedin fluorescence intensity as the linker length was increasedPotentially the weak binding sequences are held in a verticalorientation at short linker lengths which primarily exposesthe target binding site but the increased flexibility at higherlinker lengths allows for a broader range of orientations orinterintramolecular and surface interactions which cannotbe compensated for because of the poor binding affinity Non-specific binding between the target and aptamer backbone orsurface may also dominate at short linker lengths This trendheld true across all eight concentrations tested examples of25 nM and 200 nM are included as well as a 1times1M featurearray (Figures S1-3) A nonbinding streptavidin aptamermutant (SA) showed constant low background fluorescenceintensity across all linker lengths (Figure 2) The bindingof IgE to the negative mutants was always higher than SAbinding demonstrating that protein binding was reduced butnot completely eliminated

32 Linker Identity Testing Using IgE The effect of the linkeridentity was also examined as a function of linker lengthfor the aptamer IgE 17-4 by implementing A C G and Tnucleobase linkers Figure 3 shows the results of changing

Fluo

resc

ence

inte

nsity

AC

GT

0

100

200

300

400

600

500

50 15 2010

Linker length

Figure 3 Response of IgE D17-4 aptamer to 100 nM Cy3-IgE usingdifferent nucleobase linkers in PBSMTB buffer Error bars representSEM of raw fluorescence intensity values for six replicates of eachsequence

the base identity (A C G and T linker) on fluorescenceintensity at 100 nMCy3-IgE T and C bases perform similarlyat short linker lengths but T outperforms C as the lengthincreases The trend holds true for 25 nM and 200 nM IgEconcentrations as well and at 100 nM Cy3-IgE on a 1times1Mfeature array (Figures S4-6) A base demonstrates lowerfluorescence intensity than T at short lengths and then beginsto outperformT at longer lengths while G consistently showsthe lowest fluorescence intensity This result is reasonableconsidering that the C and G bases have potential to formmore complex structures potentially reducing the expecteddistance from the surface or altering the binding propertiesof the sequence In summary the results show that differentbases will provide varying signals and the T base is aconsistently high performer but may be outperformed by Abase at longer linker lengths

33 Effect of Linker Length and Identity on IgE AptamerTargetAffinity The next question investigated was how linkerlength and identity will affect the actual aptamertarget bind-ing affinity measured by dissociation constant (119870

119889) Further

information onmethods and discussion related to generationand processing of the binding curves can be found in theSupporting Information In the current study the aptamerresponse was analyzed for each linker length and identityat Cy3-IgE concentrations up to 200 nM The fluorescenceresponse increased with an increase in linker length forthe original IgE 17-4 aptamer and the positive mutants butwas close to zero for negative controls after backgroundsubtraction indicating nonspecific binding for the negativesequences (Figure S7) The microarray results suggested thatthe affinity was not significantly different between T linker(119870119889= 663 plusmn 31 nM) and A linker (119870

119889= 637 plusmn 17 nM)

or C linker (618 plusmn 38 nM) but 119870119889= 865 plusmn 81 nM for

G is statistically significant from T (119901 = 00239 95 CI)Binding constants were also determined for the mutants SM1


R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(a)

R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(b)

Figure 4 Comparison of coefficient of determination (1198772) versus linker length for (a) direct labelingmethod and (b) indirect labelingmethodCurveswere fit to a one site specific bindingwithHill slopemodel usingGraphPadPrismA (circle) C (square) G (triangle) andT (downwardtriangle) are the linkers used for IgE 17-4 aptamer The remaining sequences (SM1 DM1 37 SA and SM21) all possessed T linkers

(799plusmn38 nM) andDM1 37 (823plusmn65 nM) both of which aresignificantly (119901 = 0020 and 119901 = 00497 95 CI) higher thanIgE 17-4 T linker (Figure S7) The affinity of the interactionwas not improved by increasing the linker length despitethe gains in sensitivity from the higher overall fluorescenceintensity

The average 119870119889for the IgE 17-4 T linker (663 plusmn 31 nM)

differs from three separate binding assays where the aptamerwas not immobilized to a surface nitrocellulose (119870

119889=

9 nM) surface plasmon resonance (SPR 119870119889= 47 nM) and

fluorescence anisotropy (119870119889= 15 plusmn 4 nM) [11 20] This

is not unexpected since past investigations have generallyshown that surface immobilization of an aptamer selectedin free solution may result in diminished affinity for thetarget [10 17 18] However the microarray affinity from thecurrent study does agree with an SPR-based study where theIgE aptamer was immobilized on the surface demonstrating119870119889= 748 nM [21] The affinity reported in the current study

also varies from prior microarray studies estimating 119870119889gt

100 nM [11 19] We suspect that the differences in dye-to-protein (DP) ratios incubation time and temperature bufferconditions and background subtraction data processing stepcontribute to this discrepancy between microarray studies[11 19]

34 Effect of Linker Length and Identity on Fit Quality of IgEAptamerTarget119870

119889Curves Analyzing the binding affinity of

the aptamertarget interaction in relation to the quality ofthe fit of the data to the binding model provides additionalinformation on optimal conditions At short linker lengthsthe binding curves did not fit the data set or the error washigh in the 119870

119889estimate A plot of the coefficients of deter-

mination (1198772) across linker lengths shows the improvement

in fit as linker length increases (Figure 4(a)) All of the IgEbinders have poor data fits at short lengths and then improveuntil a plateau is reached close to 1198772 = 10 This effect mayresult from a strong dependence on nonspecific binding atshort lengths andor because the rigidity of the linkage tothe surface does not allow for the aptamer to accommodatetarget binding This implies that a minimal linker length (sim10 nucleobases) should be applied to achieve high-qualitybinding data Increasing the length of the linker past thispoint will increase the sensitivity (fluorescence intensity) ofthe response by generating higher overall signal but will notaffect binding Nonbinding sequences (SA and SM21) eitherwere not able to produce a curve fit at all (1198772 = 0) orproduced a poor curve fit which never reached acceptable(1198772 gt 095) levels As a whole the binding studies showthat both the linker length and identity can influence aptamerbinding under the conditions of this study

35 Effect of Labeling Method on IgE AptamerTarget BindingThe method used to report a binding event from the proteintarget is another important facet of microarray studiesMost dye systems used to label proteins for microarraywork possess conjugated 120587-systems reported to interact withDNA directly which may manifest in a nonspecific bindingcomponent [14] A feasible method of biotinylating the pro-tein followed by a short dye-labeled streptavidin incubationhas been shown to demonstrate less nonspecific bindingcompared to a direct dye labeling [14 16] We hypothesizedthat the lower background expected in this ldquoindirectrdquo proteinlabeling strategy may increase the signal to noise ratio (SN)and thereby lower limits of detection (LOD) compared tothe direct labelingmethodThe average fluorescence intensityof the T

10IgE 17-4 aptamer was divided by the average

Journal of Nucleic Acids 7SN

[IgE] (nM)0 50 100 150 200 250

10nM Cy3-SA (indirect)10nM Cy3-IgE (direct)

0

500

1000

1500

2000

2500

Figure 5 Signal-to-noise (SN) comparison using direct and indi-rect dye labeling under different conditions for T

10IgE 17-4 Biotin-

IgE and Cy3-IgE concentrations were 10 nM

fluorescence intensity for the SA to determine the signal tonoise ratio for each set of conditions The SN of the indirectmethod is a maximum of over 650X higher than the directlabeling method (Figure 5) At [biotin-IgE] = 10 nM [Cy3-SA] = 10 nM the LOD was improved to 4 pM from 500 pMwith [Cy3-IgE] = 10 nM in the direct method Comparisonof the images of the arrays with only 10 nM Cy3-streptavidinor 10 nM biotin-IgE + 10 nM Cy3-streptavidin shows thatthe fluorescence is a result of IgE binding rather than anonspecific interaction with streptavidin or Cy3 (Figure S8)These results show that the indirect labeling method hasenhanced signal and lower nonspecific binding compared tothe direct labeling method

36 Effect of Indirect Labeling Method on IgE AptamerTargetAffinity and 119870

119889Curve Fit Binding curves were also gener-

ated for IgE sequences using the indirect labeling methodwith [biotin-IgE] = 10 nM [Cy3-SA] = 10 nM (Figure S9)The 119870

119889for IgE 17-4 T linker was 295 plusmn 07 nM which was

approximately half that reported using the direct labelingmethod All other nucleobase linkers were close in affinityto the T linker with A linker 119870

119889= 309 plusmn 04 nM C

linker 119870119889= 308 plusmn 01 nM and G linker 119870

119889= 295 plusmn

08 nM The fact that the G linker displayed weaker affinitythan the other nucleobase linkers using the direct labelingmethod but all linkers were similar using the indirect labelingmethod may be due to minimization of the fluorescencereporter interaction with the surface using the indirectlabeling method Furthermore Cy3 may occlude the bindingsite of IgE using the direct labeling method where thebiotinylation performed for the indirect labeling method didnot interfere with binding The affinities of the IgE SM1 andIgE DM1 37 sequences also aligned with119870

119889= 288 plusmn 06 nM

and 119870119889= 299 plusmn 08 nM respectively Additionally binding

was observed for aptamer IgE 17-1 (Figure S9I 119870119889= 734 plusmn

51 nM) using the indirect labeling method where it was notevident during direct labeling (Figure S7I) This agrees with

Fluo

resc

ence

inte

nsity

DP = 028

DP = 041

DP = 24

50 1510

Linker length

0

1000

2000

3000

4000

Figure 6 Variation of dye-to-protein (DP) ratio of 100 nM Cy3-IgE for direct labeling method Error bars represent SEM for sixreplicates of IgE 17-4 aptamer with T linker

the solution-based nitrocellulose assay performed followingthe original selection for IgE 17-1 reporting a 119870

119889= 82 nM

[20] The trends of coefficient of determination are differentfrom those of the direct binding method (Figure 4(a)) in thatthere does not appear to be a strong dependence on linkerlength for the indirect labeling method (Figure 4(b)) All ofthe IgE binders are close to 1198772 = 10 at all linker lengthswhile the nonbinder fits are more variable The model couldnot accurately fit SA or SM21 binding data but it appears thatSM21 has some weak binding activity likely still in the linearrange (Figure S9H inset) Using the indirect labelingmethodresulted in higher affinity binding with less dependence onlinker identity and length than direct labeling under theconditions of this experiment

37 Effect of Labeling Method on SN for IgE AptamerTargetBinding The DP of the direct labeling method was inves-tigated next to deduce whether the cause of the enhancedbinding in the indirect method was due to the labeling of IgEPrevious work has shown that an excess of dye can occludethe binding site of the target or diminish fluorescence yieldby resonance energy transfer [9 22]Therefore the DP = 24used in the experiments described above was compared toDP = 041 and 028 At 100 nM IgE concentration Figure 6shows that reducing the DP from 24 to 041 providesthe highest fluorescence intensity of the ratios tested whilefurther reduction to 028 shows the lowest fluorescenceintensity This demonstrates the tradeoff between decreasingDP to allow for more binding and the lower fluorescenceintensities that result at some cutoff point In this experimentthe SN values were 186 for DP = 041 53 for DP = 24and 49 for DP = 028 versus 875 for 100 nM biotin-IgE with10 nM Cy3-SA Further optimization of the DP or biotin-to-protein ratio may improve the SN for the direct andindirect labeling methods respectively This shows that SNcan be improved by adjusting DP on a protein target using

8 Journal of Nucleic AcidsFl

uore

scen

ce in

tens

ity

Linker length0 10 20 30 40 50

TFBS-ATFBS-C

TFBS-GTFBS-T

0

50000

100000

150000

(a)

Linker length0 10 20 30

THBS-TRF2RF3

SAIgEATP 2542

Fluo

resc

ence

inte

nsity

0

50000

100000

150000

(b)

Figure 7 Binding of Cy3-thrombin to various sequences (a) Binding of Cy3-thrombin to TFBS aptamer using different nucleobase linkersand lengths (b) Binding of Cy3-thrombin to T linker variants of THBS (thrombin binding sequence and reported G-quartet former) RF2(riboflavin aptamer and reported G-quartet former) RF3 (riboflavin aptamer and reported G-quartet former) SA (streptavidin aptamermutant and non-G-quartet former) IgE D17-4 (IgE aptamer and non-G-quartet former) and ATP 2542 (ATP aptamer initially reported toform G-quartet but later determined to form pseudoknot structure) Error bars represent SEM of raw fluorescence intensity values for 5ndash15replicates of each sequence Linker identity was T unless otherwise specified

a direct detection method but it is still not as sensitive as theindirect labelingmethod Binding of aptamer IgE17 1was alsorestored at DP = 041 where it was not observed at DP =24 illustrating that the binding properties of an aptamer areinfluenced by DP

38 Effect of Linker Length and Identity for Cy3-ThrombinBinding G-Quartet Aptamers We next examined the inter-action of Cy3-thrombin with aptamers reported to form G-quartet structures The binding of Cy3-thrombin with the15-mer thrombin aptamer TFBS the 29-mer thrombin G-quartet aptamer THBS the 2-tiered G-quartet riboflavinaptamer RF2 and the 3-tiered G-quartet riboflavin aptamerRF3 was investigated [23ndash25] The linker length and identityalso affect thrombin binding to TFBS though in a differentmanner than IgE binding IgE 17-4 (Figure 7(a)) Similar tothe IgE profile T linker shows a consistent fluorescenceincrease to long linker lengths and G linker is regularlythe worst performer across all linker lengths Howeverfluorescence signal actually decreases with increasing lengthfor A linker where it increased for all base linkers inIgEIgE 17-4 binding A is initially the best performer forTFBSthrombin and then decreases at longer lengths incontrast to IgEIgE 17-4 results while C results are variableStructurally this may indicate that certain bases included inthe linker may interact with the aptamer sequence and affecttarget binding This shows that the performance will varywith the linker ID depending on the aptamer target pair andthe effects should be investigated prior to implementationThe effects of thrombin binding THBS (Figure 7(b)) aremuchmore dramatic across linker lengths as the fluorescenceintensity increases from slightly above background at T

0

to two orders of magnitude increase at T15 This agrees

with previous reports that contend that the THBS aptamerconsists of a structure that requires extra spatial mobility tofold into an active conformation due to the longer duplexbetween the 51015840 and 31015840 terminus and has reduced stabilityin the G-quartet compared to TFBS [24] Also thrombindemonstrates enhanced binding to RF3 compared to TFBSbut no binding to the 2-tiered RF2 aptamer (Figure 7(b))Binding of thrombin to RF3 is not unexpected due to studiesthat showed that the core G-quartet region of TFBS (similarin RF3) is necessary for binding while changes to theidentities of the loop sequences are tolerated with differentlevels of success [16 26 27] The lack of RF2 binding isexpected as well due to reports indicating that RF2 is lessstable than RF3 in a Na+ based buffer and requires a higherK+ content for structural stability to fold into the G-quartetstructure [25] Thrombin did not bind significantly to SAIgE 17-4 or an ATP aptamer (ATP 2542) initially reportedto fold into a G-quartet but later shown to actually form apseudoknot instead (Figure 7(b) [28 29])

39 Effect of Linker Length and Identity for Cyanine DyeInteraction with G-Quartet Aptamers Previous research hasshown that fluorescent dyes can intercalate between G-quartets [30ndash32]This property was exploited to propose thatmicroarrays can be used to determine whether a sequencefolds into a G-quartet structure andmay also be used to studystructural properties of G-quartet aptamers under differentconditions by adding dye directly to the array Figure 8shows a strong response for TFBS with Cy3 dye whichvaries depending on linker identity and length Contrary tothe interaction between Cy3-thrombin and TFBS G linker


uore

scen

ce in

tens

ity


TFBS-ATFBS-CTFBS-GTFBS-TTHBS-T

RF2RF3SAIgEATP 2542

0

10000

20000

30000

Figure 8 Binding of Cy3 dye in PBSMTB buffer to a variety of G-quartet and non-G-quartet aptamer sequences Error bars representSEM of raw fluorescence intensity values for 5ndash15 replicates of eachsequence Linker identity was T unless otherwise specified

increases in intensity as linker length increases presumablydue to more G-quartets forming within the linker C is theworst performer at long lengths potentially due to the Cbases pairing with some of the Grsquos in the G-quartet aptamerportion altering the binding conformation of TFBS A linkerdemonstrates a trend similar to that of thrombin binding andT is the best performer increasing and then plateauing Thestronger dependence of Cy3 binding TFBS as T linker lengthis increased as compared to thrombin binding may be due tothe increased space required to compensate for electrostaticrepulsion of the sulfate groups of Cy3 by the DNA backboneThe larger size of thrombinmay also sterically limit the bene-fits of extending the aptamer from the array surface AptamerRF3 displayed enhanced fluorescence when interacting withCy3 dye while RF2 did not demonstrate fluorescence abovebackground similar to the interaction with thrombin THBSfluorescencewas slightly above SAbackgroundbut drasticallydiminished compared to the THBSthrombin interactionThis may be due to competition between K+ and Cy3 forbinding or by K+ inducing a conformational change whichdiminishes Cy3 binding Cy5 dye also interacts with TFBSfollowing a trend very similar to Cy3 binding in PBSMTB(Figure S10) This work indicates that cyanine dyes can beused to probe whether aptamers fold into a G-quartet ona microarray surface and to glean structural informationrelated to the binding parameters Based on the performanceof all the aptamertarget pairs studied in this work T base isrecommended as a safe starting point for sensor developmentapplications

310 Effect of Buffer Conditions on Cyanine Dye Interactionwith G-Quartet Aptamers In order to further probe thestructural features of the G-quartet aptamers the sequences

were exposed to various buffer conditions with Cy3 targetFigure 8 illustrated that RF2 did not bind Cy3 dye inPBSMTB buffer which was postulated to be due to RF2decreased binding in Na+ buffers and high dependence onK+ ion The same experiment was then repeated to assessRF2 binding in the original selection buffer containing noNa+ and high (300mM) K+ RF2 binding was equivalentto RF3 binding under these conditions indicating thatthe RF buffer allowed RF2 to form a properly folded G-quartet structure (Figure 9(a)) However RF3 binding wassignificantly diminished in the original RF buffer comparedto PBS-based buffer (Figure 7(b)) where it demonstratedhigher fluorescence intensity than TFBS THBS binding wasalso enhanced in the Cy3-targeted RF buffer compared toPBSMTB (Figure 9(a)) This led to questions on the role ofthe K+ ion for these different G-quartet aptamers bindingCy3dye When PBS buffer was used without any K+ componentRF2 binding was eliminated and the binding of all probeshad higher error for each measurement (Figure 9(b)) TFBSbinding was similar between RF and PBS (no K+) bufferswhile THBS and RF3 binding were enhanced in the PBS(no K+) buffer indicating a minimal dependence on K+ forG-quartet binding under the experimental conditions Thisshows that certain sequences have different sensitivities tobuffer composition and some aptamers may show enhancedbinding in buffers other than the original selection bufferMicroarrays afford the researcher the opportunity to rapidlystudy various conditions in parallel to optimize the perfor-mance for a biosensor application

4 Conclusions

The current research investigated the binding and structuralproperties of aptamers immobilized on a microarray surfacewhich were initially selected by solution-based SELEX Themicroarray format was specifically chosen as a platformwhich would allow for the study of aptamers in an immo-bilized format similar to most biosensing applications Asa result of this study we were able to make generalizationsregarding conditions researchers may wish to use as astarting point for future microarray aptamer studies This isparticularly interesting because most microarray protocolsprovided by commercial vendors are for cDNA hybridiza-tions which are much stronger than the noncovalent inter-actions experienced by aptamertarget binding Thereforeto date most aptamer microarray procedures are optimizedfrom scratch for each aptamer target pair In the currentstudy we show that careful attention should be paid to theidentity of the linker used as well as to the distance ofthe aptamer from the microarray surface (linker length) Tlinker appears to be a consistent choice for linker acrossdifferent types of aptamertarget pairs while G linker wasthe weakest performer perhaps due to the formation ofintralinker secondary structure as the length increases Wealso show that a minimal linker length may be required toproduce reliable binding data and that extending the lengthpast approximately 10 nucleobases increases the sensitivity ofthe response but does not affect the affinity Furthermorean indirect labeling method was shown to have a higher


uore

scen

ce in

tens

ity


TFBS-ATFBS-CTFBS-GTFBS-TTHBS

RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)

Figure 9 Binding of Cy3 dye to a variety of G-quartet and non-G-quartet aptamer sequences using (a) initial RF aptamer selection buffer(b) PBSMTB without K+ Error bars represent SEM of raw fluorescence intensity values for 5ndash15 replicates of each sequence Linker identitywas T unless otherwise specified

SN than the direct labeling method potentially due in partto minimized interaction of the fluorescent reporter withthe microarray surface The SN can be modulated to somedegree by changing the DP of the dye linked with target butdid not reach the high levels attained by the indirect labelingmethod Collectively this work describes some baselineconditions which may reduce the optimization time of futurestudies A further result of these studies was to report novelinformation on specific aptamertarget interactions As anexample the indirect binding method demonstrated bindingfor the weak binding IgE 17-1 and ID17-4SM21 sequenceswhich was not observed using the direct binding methodThis work also experimentally indicated that theoreticalpropositions of thrombin and riboflavin aptamer structuresmay be accurate under different conditions In addition weshow that microarrays can be used as a simple method fordetermining whether an aptamer forms a G-quartet structureby adding a compatible dye to the surface This methodis simpler than previous methods that require sequence-specific assignments to determine aptamer structure and iscomplementary to the recently reported benefits of structuralscreening by a combination of imino proton NMR and 2DNOESY [33]

The simplicity of the microarray experiments and rapidtimeframe from lab work to data analysis promote the useof the technology for screening the effect of a variety ofconditions on aptamer binding to optimize the interaction Inaddition to those assayed in this work the microarray couldbe instrumental in rapidly assessing aptamer candidates forfunction at enhanced temperature such as those mimickingphysiological conditions Microarray technology could alsobe applied to rapidly assess a multitude of aptamers for

cross-reactivity to physiologically relevant targets as would beanticipated to have importance in point-of-care applicationsThis work demonstrated design considerations and benefitsofmicroarray technology toward an improved understandingof the aptamer structurefunction relationship aimed at effec-tively integrating solution-selected aptamers onto biosensingplatforms

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors thank Dr Richard Chapleau for assistance with119870119889data interpretation and Agilent Technologies for the loan

of the Agilent High-Resolution Microarray Scanner Thisresearch was performed while the author Jennifer A Martinheld a National Research Council Research AssociateshipAward at Wright-Patterson Air Force Base This work wassupported by the Air Force Office of Scientific Research AirForce Human Signatures Branch (711 HPWRHXB) CoreAir Force Research Laboratory Bio-X Strategic TechnologyThrust Biotronics and Defense Forensic Science Center

References

[1] R Stoltenburg C Reinemann and B Strehlitz ldquoSELEXmdasha(r)evolutionary method to generate high-affinity nucleic acidligandsrdquo Biomolecular Engineering vol 24 no 4 pp 381ndash4032007


[2] E Luzi M Minunni S Tombelli and M Mascini ldquoNew trendsin affinity sensing aptamers for ligand bindingrdquo TrACmdashTrendsin Analytical Chemistry vol 22 no 11 pp 810ndash818 2003

[3] T S Romig C Bell and D W Drolet ldquoAptamer affinitychromatography combinatorial chemistry applied to proteinpurificationrdquo Journal of Chromatography B Biomedical Sciencesand Applications vol 731 no 2 pp 275ndash284 1999

[4] M Blank T Weinschenk M Priemer and H SchluesenerldquoSystematic evolution of a DNA aptamer binding to rat braintumor microvesselsrdquo The Journal of Biological Chemistry vol276 no 19 pp 16464ndash16468 2001

[5] Y Wu K Sefah H Liu R Wang and W Tan ldquoDNA aptamer-micelle as an efficient detectiondelivery vehicle toward cancercellsrdquo Proceedings of the National Academy of Sciences of theUnited States of America vol 107 no 1 pp 5ndash10 2010

[6] M Famulok J S Hartig and G Mayer ldquoFunctional aptamersand aptazymes in biotechnology diagnostics and therapyrdquoChemical Reviews vol 107 no 9 pp 3715ndash3743 2007

[7] A D Ellington and J W Szostak ldquoIn vitro selection of RNAmolecules that bind specific ligandsrdquo Nature vol 346 no 6287pp 818ndash822 1990

[8] C Tuerk and L Gold ldquoSystematic evolution of ligands byexponential enrichment RNA ligands to bacteriophage T4DNApolymeraserdquo Science vol 249 no 4968 pp 505ndash510 1990

[9] J R Collett J C Eun and A D Ellington ldquoProduction andprocessing of aptamer microarraysrdquo Methods vol 37 no 1 pp4ndash15 2005

[10] E J Cho J R Collett A E Szafranska and A D EllingtonldquoOptimization of aptamer microarray technology for multipleprotein targetsrdquo Analytica Chimica Acta vol 564 no 1 pp 82ndash90 2006

[11] E Katilius C Flores and N W Woodbury ldquoExploring thesequence space of a DNA aptamer using microarraysrdquo NucleicAcids Research vol 35 no 22 pp 7626ndash7635 2007

[12] J-G Walter O Kokpinar K Friehs F Stahl and T ScheperldquoSystematic investigation of optimal aptamer immobilizationfor protein-microarray applicationsrdquo Analytical Chemistry vol80 no 19 pp 7372ndash7378 2008

[13] O Srivannavit M Gulari Z Hua et al ldquoMicrofluidic reactorarray device for massively parallel in situ synthesis of oligonu-cleotidesrdquo Sensors and Actuators B Chemical vol 140 no 2 pp473ndash481 2009

[14] M Platt W Rowe D C Wedge D B Kell J Knowles andP J R Day ldquoAptamer evolution for array-based diagnosticsrdquoAnalytical Biochemistry vol 390 no 2 pp 203ndash205 2009

[15] C Bock M Coleman B Collins et al ldquoPhotoaptamer arraysapplied to multiplexed proteomic analysisrdquo Proteomics vol 4no 3 pp 609ndash618 2004

[16] M Platt W Rowe J Knowles P J Day and D B Kell ldquoAnalysisof aptamer sequence activity relationshipsrdquo Integrative Biologyvol 1 no 1 pp 116ndash122 2009

[17] Y-H Lao K Peck and L-C Chen ldquoEnhancement of aptamermicroarray sensitivity through spacer optimization and avidityeffectrdquo Analytical Chemistry vol 81 no 5 pp 1747ndash1754 2009

[18] J R Collett E J Cho J F Lee et al ldquoFunctional RNAmicroar-rays for high-throughput screening of antiprotein aptamersrdquoAnalytical Biochemistry vol 338 no 1 pp 113ndash123 2005

[19] N O Fischer J B-H Tok and T M Tarasow ldquoMassively paral-lel interrogation of aptamer sequence structure and functionrdquoPLoS ONE vol 3 no 7 Article ID e2720 2008

[20] T W Wiegand P B Williams S C Dreskin M Jouvin JKinet and D Tasset ldquoHigh-affinity oligonucleotide ligands tohuman IgE inhibit binding to Fc120576 receptor Irdquo The Journal ofImmunology vol 157 no 1 pp 221ndash230 1996

[21] J Wang R Lv J Xu D Xu and H Chen ldquoCharacterizing theinteraction between aptamers and human IgE by use of surfaceplasmon resonancerdquo Analytical and Bioanalytical Chemistryvol 390 no 4 pp 1059ndash1065 2008

[22] C D Hahn C K Riener and H J Gruber ldquoLabeling ofantibodies with Cy3- Cy35- Cy5- and Cy55-monofunctionaldyes at defined dyeprotein ratiosrdquo Single Molecules vol 2 no2 pp 149ndash154 2000

[23] L C Bock L C Griffin J A Latham E H Vermaas and J JToole ldquoSelection of single-stranded DNA molecules that bindand inhibit human thrombinrdquo Nature vol 355 no 6360 pp564ndash566 1992

[24] D M Tasset M F Kubik and W Steiner ldquoOligonucleotideinhibitors of human thrombin that bind distinct epitopesrdquoJournal of Molecular Biology vol 272 no 5 pp 688ndash698 1997

[25] C T Lauhon and J W Szostak ldquoRNA aptamers that bind flavinand nicotinamide redox cofactorsrdquo Journal of the AmericanChemical Society vol 117 no 4 pp 1246ndash1257 1995

[26] A Bugaut and S Balasubramanian ldquoA sequence-independentstudy of the influence of short loop lengths on the stability andtopology of intramolecular DNA G-quadraplexesrdquo Biochem-istry vol 47 no 2 pp 689ndash697 2008

[27] I Smirnov and R H Shafer ldquoEffect of loop sequence and sizeonDNAaptamer stabilityrdquoBiochemistry vol 39 no 6 pp 1462ndash1468 2000

[28] D E Huizenga and J W Szostak ldquoA DNA aptamer that bindsadenosine and ATPrdquo Biochemistry vol 34 no 2 pp 656ndash6651995

[29] C H Lin and D J Patel ldquoStructural basis of DNA foldingand recognition in an AMP-DNA aptamer complex distinctarchitectures but common recognition motifs for DNA andRNA aptamers complexed to AMPrdquo Chemistry amp Biology vol4 no 11 pp 817ndash832 1997

[30] Y Li C R Geyer and D Sen ldquoRecognition of anionicporphyrins by DNA aptamersrdquo Biochemistry vol 35 no 21 pp6911ndash6922 1996

[31] CWilson and JW Szostak ldquoIsolation of a fluorophore-specificDNA aptamer with weak redox activityrdquo Chemistry amp Biologyvol 5 no 11 pp 609ndash617 1998

[32] V BKovalskaMY Losytskyy SMYarmoluk I Lubitz andAB Kotlyar ldquoMono and trimethine cyanines Cyan 40 and Cyan2 as probes for highly selective fluorescent detection of non-canonical DNA structuresrdquo Journal of Fluorescence vol 21 no1 pp 223ndash230 2011

[33] J A Martin P A Mirau Y Chushak et al ldquoSingle-roundpatterned DNA library microarray aptamer lead identificationrdquoJournal of Analytical Methods in Chemistry vol 2015 Article ID137489 8 pages 2015

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research



Microbiology


Microarray technology can be used to probe the func-tional characteristics of aptamers selected by solution-basedSELEX in an immobilized format mimicking the ultimatebiosensing application Microarrays indicate target bindingby identifying locations of fluorescence from a fluorescentlylabeled target and correlating them with the position ofknown sequences on the array Early attempts in microarraydesign suffered from challenges related to oligonucleotidedeposition by spotting presynthesized probes onto an acti-vated array surface specifically in reproducibility whenconditions (humidity buffer etc) varied from batch to batch[12 13] Oligonucleotide spotting also rendered testing ofthousands of sequences cost prohibitive due to the expenseof synthesizing each probe individually prior to depositiononto the array Current microarray technology based onsynthesis from the array surface rather than oligonucleotidespotting produces high reproducibility and allows for up to106 completely custom 80 mer DNA sequences synthesizedper array The massively parallel nature of microarray testingdrastically reduces the time and cost investments associatedwith assaying the functional characteristics of aptamerscompared to traditionalmethods of individually synthesizingand purifying each sequence of interest For example amicroarray experiment can be completed in less than oneday with data analysis requiring another 1-2 days A singlemillion feature array (Agilent 1times1M) similar to those usedin the current study at a price of sim$600 can assay up toa million sequences simultaneously compared to sim$135 forone sequence HPLC purified from commercial vendors fortraditional analysis

Onemajor challenge associated with applyingmicroarraytechnology to aptamer studies lies in the fact that current pro-tocols have been mostly developed for gene expression stud-ies that function by direct hybridization of the fluorescent-tagged genomic fragments to cDNA on the microarray Thistype of interaction forms an extremely stable duplex betweenthe base pairs of the two strands whereas aptamerproteintarget interactions may heavily rely on one or a fewnoncovalent interactions For example the binding of theimmunoglobulin E (IgE) aptamer has been shown to be com-pletely destroyed following change of a single base within theaptamer sequence [11] Moreover the majority of dye systemsused to label protein targets (ldquodirect labelingrdquo) formicroarraywork possess conjugated 120587-systems reported to interact withDNA directly [14] This observation renders it challengingto differentiate between sequences which are binding theactual protein target and those that instead interact with thedye Gold and coworkers reported using photoactive basesubstitutions to covalently cross-link proteins to the arraysfollowed by stringent washing steps to remove nonspecificbinders [15] However commercial microarray manufactur-ers do not currently offer these base substitutions in theirarray synthesis portfolio A feasible method of biotinylatingthe protein followed by a short dye-labeled streptavidinincubation (indirect labeling) demonstrated less nonspecificbinding compared to direct dye labeling [14 16] Further-more many parameters will affect the aptamer response thatmay not be as significant in cDNA hybridizations including

probe density proximity of the binding site to the surfaceimmobilization orientation (51015840 or 31015840 end) and the bufferused in the studies [10 17]Therefore protocols and methodsfor aptamers must be developed and optimized by theindividual researcher accounting for challenges specific to theapplication

As is the case with biosensor applications the responseof an aptamer in solution can be significantly different thanwhen it is attached to a microarray surface [9] Severalstudies report differences in aptamer affinity when com-paring microarray and solution-based dissociation constantmeasurements [10 11] In most cases the affinities of theseaptamers were significantly lower on the microarray thanwhen tested in the solution-based conditions used duringtheir selection [18] Significant work has been performed bydifferent groups in the use of aptamers for target detection incustom-mademicroarray settings It has been confirmed thatknown DNA aptamers can be immobilized on a microarrayand conditions to preserve their activity can be found andthe conditions used for these studies should mimic theconditions used during the initial solution-based selectionprocess [12 18] Prior work has also demonstrated thatextending the distance of the aptamers from the arraysurface enhances target binding fluorescence intensity isproportional to the concentration of target added and higherrelative fluorescence values between probes on the samearray indicate higher affinity interactions [11 16] Substan-tial work has also been presented describing the ability ofmicroarrays to assay target binding after aptamer mutationstruncations or changing conditions but these studies haveeither not analyzed multiple aptamers binding multipletargets or not utilized specific conditions to analyze struc-tural and functional information regarding target binding[11 18 19]

The main goal of the current studies was to investigatemicroarrays as a model system to examine aptamer bindingin an immobilized format mimicking that of biosensorapplications During the course of these studies several dif-ferent conditions were investigated related to the functionalcharacteristics of multiple aptamers on custom microarraysThese enabled generalizations related to potential startingpoints for researchers wishing to utilize microarrays in futureaptamer studies as well as providing insights into specificaptamertarget interactions Specifically we investigated theeffect of linker length with IgE as a control and expanded thework to analyze the effect of the linker identity dissociationconstant (119870

119889) under varying conditions effect of the dye-to-

protein ratio (DP) and comparison of a direct and indirectlabeling method [11] The schematic in Figure 1 depicts asummary of the variables investigated using microarrays inthe current study In addition the linker length and identityeffects were extended to include several guanine-quartet (G-quartet) aptamers in relation to thrombin and dye bindingFurthermore we utilized the known ability of G-quartets tobind 120587-conjugated dye systems to assess the structural char-acteristics of G-quartet aptamers under different conditionsThis enables a correlation between the conditions and theaptamer structure without concern regarding whether theselected conditions may impart conformational changes on




extraction

Biotinylatedtarget


(short)


TTT

TTT

TTTT

TTTTT

TTTTTT

TTTT

AAAA

CCCC

GGGG




Data

Guanine (G)

N

N

NN

HN

H

HNN

NN

N

N

H

H

NH2

NH2

NH

OO

O

O

H2N

AATGACTCG

CTAGAGTCA

ACGGGTACT GGTACGATA

CAGAAAATC

TTTACTATAG






2HPO4






MgCl2




20 and 1 BSA


1mMMgCl2


2


2+ 01 Tween-20 + 1 BSA


















uore

scen

ce in

tens

ity


ID17-4DM13_25

ID17-4SM17

ID17-4SM21

ID17-4SM28

SA

50 15 2010

Linker length

0

100

200

300

400

500




Fluo

resc

ence

inte

nsity

AC

GT

0

100

200

300

400

600

500

50 15 2010

Linker length




119889) Further


119889= 637 plusmn 17 nM)




R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(a)

R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(b)





119889=















[IgE] (nM)0 50 100 150 200 250


0

500

1000

1500

2000

2500


10IgE 17-4 Biotin-








119889= 309 plusmn 04 nM C


119889= 295 plusmn


119889= 288 plusmn 06 nM




Fluo

resc

ence

inte

nsity

DP = 028

DP = 041

DP = 24

50 1510

Linker length

0

1000

2000

3000

4000



119889= 82 nM




uore

scen

ce in

tens

ity


TFBS-ATFBS-C

TFBS-GTFBS-T

0

50000

100000

150000

(a)


THBS-TRF2RF3

SAIgEATP 2542

Fluo

resc

ence

inte

nsity

0

50000

100000

150000

(b)




0





uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000





4 Conclusions



uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)







Acknowledgments



References










































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology




extraction

Biotinylatedtarget


(short)


TTT

TTT

TTTT

TTTTT

TTTTTT

TTTT

AAAA

CCCC

GGGG




Data

Guanine (G)

N

N

NN

HN

H

HNN

NN

N

N

H

H

NH2

NH2

NH

OO

O

O

H2N

AATGACTCG

CTAGAGTCA

ACGGGTACT GGTACGATA

CAGAAAATC

TTTACTATAG






2HPO4






MgCl2




20 and 1 BSA


1mMMgCl2


2


2+ 01 Tween-20 + 1 BSA


















uore

scen

ce in

tens

ity


ID17-4DM13_25

ID17-4SM17

ID17-4SM21

ID17-4SM28

SA

50 15 2010

Linker length

0

100

200

300

400

500




Fluo

resc

ence

inte

nsity

AC

GT

0

100

200

300

400

600

500

50 15 2010

Linker length




119889) Further


119889= 637 plusmn 17 nM)




R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(a)

R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(b)





119889=















[IgE] (nM)0 50 100 150 200 250


0

500

1000

1500

2000

2500


10IgE 17-4 Biotin-








119889= 309 plusmn 04 nM C


119889= 295 plusmn


119889= 288 plusmn 06 nM




Fluo

resc

ence

inte

nsity

DP = 028

DP = 041

DP = 24

50 1510

Linker length

0

1000

2000

3000

4000



119889= 82 nM




uore

scen

ce in

tens

ity


TFBS-ATFBS-C

TFBS-GTFBS-T

0

50000

100000

150000

(a)


THBS-TRF2RF3

SAIgEATP 2542

Fluo

resc

ence

inte

nsity

0

50000

100000

150000

(b)




0





uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000





4 Conclusions



uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)







Acknowledgments



References










































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology




MgCl2




20 and 1 BSA


1mMMgCl2


2


2+ 01 Tween-20 + 1 BSA


















uore

scen

ce in

tens

ity


ID17-4DM13_25

ID17-4SM17

ID17-4SM21

ID17-4SM28

SA

50 15 2010

Linker length

0

100

200

300

400

500




Fluo

resc

ence

inte

nsity

AC

GT

0

100

200

300

400

600

500

50 15 2010

Linker length




119889) Further


119889= 637 plusmn 17 nM)




R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(a)

R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(b)





119889=















[IgE] (nM)0 50 100 150 200 250


0

500

1000

1500

2000

2500


10IgE 17-4 Biotin-








119889= 309 plusmn 04 nM C


119889= 295 plusmn


119889= 288 plusmn 06 nM




Fluo

resc

ence

inte

nsity

DP = 028

DP = 041

DP = 24

50 1510

Linker length

0

1000

2000

3000

4000



119889= 82 nM




uore

scen

ce in

tens

ity


TFBS-ATFBS-C

TFBS-GTFBS-T

0

50000

100000

150000

(a)


THBS-TRF2RF3

SAIgEATP 2542

Fluo

resc

ence

inte

nsity

0

50000

100000

150000

(b)




0





uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000





4 Conclusions



uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)







Acknowledgments



References










































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


uore

scen

ce in

tens

ity


ID17-4DM13_25

ID17-4SM17

ID17-4SM21

ID17-4SM28

SA

50 15 2010

Linker length

0

100

200

300

400

500




Fluo

resc

ence

inte

nsity

AC

GT

0

100

200

300

400

600

500

50 15 2010

Linker length




119889) Further


119889= 637 plusmn 17 nM)




R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(a)

R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(b)





119889=















[IgE] (nM)0 50 100 150 200 250


0

500

1000

1500

2000

2500


10IgE 17-4 Biotin-








119889= 309 plusmn 04 nM C


119889= 295 plusmn


119889= 288 plusmn 06 nM




Fluo

resc

ence

inte

nsity

DP = 028

DP = 041

DP = 24

50 1510

Linker length

0

1000

2000

3000

4000



119889= 82 nM




uore

scen

ce in

tens

ity


TFBS-ATFBS-C

TFBS-GTFBS-T

0

50000

100000

150000

(a)


THBS-TRF2RF3

SAIgEATP 2542

Fluo

resc

ence

inte

nsity

0

50000

100000

150000

(b)




0





uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000





4 Conclusions



uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)







Acknowledgments



References










































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(a)

R2

ACGT

SM1

DM1_37SASM21

00

02

04

06

08

10

50 15 2010

Linker length

(b)





119889=















[IgE] (nM)0 50 100 150 200 250


0

500

1000

1500

2000

2500


10IgE 17-4 Biotin-








119889= 309 plusmn 04 nM C


119889= 295 plusmn


119889= 288 plusmn 06 nM




Fluo

resc

ence

inte

nsity

DP = 028

DP = 041

DP = 24

50 1510

Linker length

0

1000

2000

3000

4000



119889= 82 nM




uore

scen

ce in

tens

ity


TFBS-ATFBS-C

TFBS-GTFBS-T

0

50000

100000

150000

(a)


THBS-TRF2RF3

SAIgEATP 2542

Fluo

resc

ence

inte

nsity

0

50000

100000

150000

(b)




0





uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000





4 Conclusions



uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)







Acknowledgments



References










































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


[IgE] (nM)0 50 100 150 200 250


0

500

1000

1500

2000

2500


10IgE 17-4 Biotin-








119889= 309 plusmn 04 nM C


119889= 295 plusmn


119889= 288 plusmn 06 nM




Fluo

resc

ence

inte

nsity

DP = 028

DP = 041

DP = 24

50 1510

Linker length

0

1000

2000

3000

4000



119889= 82 nM




uore

scen

ce in

tens

ity


TFBS-ATFBS-C

TFBS-GTFBS-T

0

50000

100000

150000

(a)


THBS-TRF2RF3

SAIgEATP 2542

Fluo

resc

ence

inte

nsity

0

50000

100000

150000

(b)




0





uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000





4 Conclusions



uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)







Acknowledgments



References










































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


uore

scen

ce in

tens

ity


TFBS-ATFBS-C

TFBS-GTFBS-T

0

50000

100000

150000

(a)


THBS-TRF2RF3

SAIgEATP 2542

Fluo

resc

ence

inte

nsity

0

50000

100000

150000

(b)




0





uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000





4 Conclusions



uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)







Acknowledgments



References










































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000





4 Conclusions



uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)







Acknowledgments



References










































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(a)Fl

uore

scen

ce in

tens

ity



RF2RF3SAIgEATP 2542

0

10000

20000

30000

40000

50000

(b)







Acknowledgments



References










































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology









































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology








Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

Documents

Research Article Microarrays as Model Biosensor Platforms ...Research Article Microarrays as Model Biosensor Platforms to Investigate the Structure and Affinity of Aptamers JenniferA.Martin,