Intrinsically Labeled Fluorescent Oligonucleotide Probes on Quantum Dots for Transduction of Nucleic Acid

Hybridization

Anna Shahmuradyan and Ulrich J. Krull*

Chemical Sensors Group, Department of Chemical and Physical Sciences, University of Toronto

Mississauga, 3359 Mississauga Road, Mississauga ON, L5L 1C6, Canada.

*Author to whom correspondence should be addressed: ulrich.krull@utoronto.ca

1

ABSTARCT

Quantum dots (QDs) have been widely used in chemical and biosensing due to their unique

photoelectrical properties, and are well suited as donors in fluorescence resonance energy

transfer (FRET). Selective hybridization interactions of oligonucleotides on QDs have been

determined by FRET. Typically, the QD-FRET constructs have made use of labeled targets or

have implemented labeled sandwich format assays to introduce dyes in proximity to the QDs for

the FRET process. The intention of this new work is to explore a method to incorporate the

acceptor dye into the probe molecule. Thiazole orange (TO) derivatives are fluorescent

intercalating dyes that have been used for detection of double-stranded nucleic acids. One such

dye system has been reported in which single-stranded oligonucleotide probes were doubly

labeled with adjacent thiazole orange derivatives. In the absence of the fully complementary

(FC) oligonucleotide target the dyes form an H-aggregate, which results in quenching of

fluorescence emission due to excitonic interactions between the dyes. The hybridization of the

FC target to the probe provides for dissociation of the aggregate as the dyes intercalate into the

double stranded duplex, resulting in increased fluorescence. This work reports investigation of

the dependence of the ratiometric signal on the type of linkage used to conjugate the dyes to the

probe, the location of the dye along the length of the probe, and the distance between adjacent

dye molecules. The limit of detection for 34mer and 90mer targets was found to be identical, and

was 10 nM (2 pmol), similar to analogous QD-FRET using labeled oligonucleotide target. The

detection system could discriminate a one base pair mismatch (1BPM) target, and was functional

without substantial compromise of the signal in 75% serum. The 1BPM was found to reduce

background signal, indicating that the structure of the mismatch affected the environment of the

intercalating dyes.

2

Keywords: fluorescence resonance energy transfer, quantum dot, oligonucleotide, hybridization, intercalation

INTRODUCTION

Cyanine dyes have been widely used for a variety of biological and biomedical applications

involving nucleic acids, such as staining of DNA in agarose gel and capillary electrophoresis

separations1,2, detection of nucleic acids after high-performance liquid chromatography

separations3, as labels in polymerase chain reaction4 , and as acceptors in FRET5,6. The

properties that make cyanine dyes useful in such applications include high affinity for double-

stranded nucleic acids, large extinction coefficients associated with strong π-π absorption, and

quantum yield that increases by orders of magnitude upon intercalation into double-stranded

DNA7.

Cyanine dyes can form aggregates in aqueous solutions due to π stacking interactions that are

associated with the polarizability and hydrophobicity of the dyes8. In such aggregates the

fluorescence is quenched due to excitonic interactions of the transition dipoles. This optical

property of the aggregates offers potential for development of DNA detection bioassays based on

an “on-off” switching system as reported by Okamoto et al.9,10. Single-stranded oligonucleotide

probes were labeled with two adjacent derivatives of thiazole orange dye that formed an H-

aggregate dimer in aqueous solutions, resulting in quenching of the fluorescence emission of the

dye molecules. Upon hybridization of the probe with complementary target, the dimer

dissociated interrupting the excitionic interactions and resulting in enhancement of fluorescence

emission due to intercalation of the dye molecules into the hybrid structure9,10. This approach to

transduction of hybridization makes use of intrinsically labeled oligonucleotide probes that allow

3

reversible determination of unlabeled target in a single step, and eliminates need for additional

reagents such as found in sandwich assays or in staining of hybrids using intercalators.

Figure 1. Schematic representation of the dye labeled probe on the surface of a QD. The dye is excited by FRET with the QD serving as donor. (a) In the absence of the target the dye molecules form an H-aggregate, which results in suppression of fluorescence emission due to excitonic interaction. (b) The hybridization of the target results in the disruption of the excitonic

interaction and strong emission from intercalated dye.

The work reported herein explores the analytical performance of the general configuration shown

in Figure 1, where the concept of use of intrinsically labeled probes is combined with optical

excitation by quantum dots (QDs). The DNA probes are immobilized on the surface of QDs and

the fluorescent dyes are excited via Fluorescent Resonance Energy Transfer (FRET) from the

QD donors. This configuration offers several advantages to the analytical performance. The QDs

have broad absorption and narrow emission bands, so that a range of absorption wavelengths can

be easily converted into a well-defined emission for the FRET process. FRET-based detection

provides inherent spatial selectivity as FRET efficiency is distance dependent. The FRET

4

excitation interrogates a distance on the scale of 10 nm (Equation 1), so that only the surface

interactions on the QDs are sampled.

E=a Ro

6

r 6+a Ro6 Equation 1

where E is the FRET efficiency, r is the distance between the donor and the acceptor, a is the

total number of the acceptors and Ro is the Förster distance, the distance at which the energy

transfer has 50% efficiency11. The interdependence of QD and dye emission allows for

ratiometric detection and opportunities for improved efficiency of hybridization based on the

high curvature of the nanoparticle surface, multiplexed analysis using concurrent spectroscopic

color channels and signal amplification by proximity of an acceptor to multiple donors in solid-

state assay formats.

Ratiometric detection is based on the calculation of the emission intensity ratio of the donor and

the acceptor, and offers improved precision while accounting for donor photoluminescence (PL)

quenching and sensitization of acceptor PL12. This investigation of the FRET-based transduction

system considers a variety of structural motifs associated with the location of the cyanine dyes to

extend the original work of Okamoto et al.9,10. Permutations include investigation of placement

of the dyes on nucleobases, on the phosphate backbone, and the effect of the distance of

separation of dyes.

EXPERIMENTAL SECTION

A detailed description of the experimental procedures, instrumentation and data analysis can be

found in the Supporting Information.

5

Reagents and oligonucleotides

Qdot® 525 ITK™ Streptavidin Conjugate Kits were from Life Technologies, a ThermoFisher

brand. Oligonucleotides were from Integrated DNA Technologies (Coralville, IA) and were

purified by either standard desalting or HPLC by the manufacturer. The oligonucleotide

sequences were dissolved in autoclaved MilliQ water (purified water from a Milli-Q cartridge

filtration system with a resistivity of 18.2 MΩ.cm) and stored at −20 °C. 2-methylbenzoxazole

(C8H7NO, 99%), 5-bromovaleric acid (Br(CH2)4COOH, 97%), 4-dimethylamino benzaldehyde

((CH3)2NC6H4CHO, ≥99.0%, HPLC), acetic anhydride ((CH3CO)2O, ReagentPlus®, ≥99%) N-

hydroxysuccinimide (NHS, C4H5NO3, 98%), 1,2- dichlorobenzene (C6H4Cl2, anhydrous, 99%) N-

(3-Dimethylaminopropyl)-N′-ethylcarbodiimide hydrochloride (EDC, C8H17N3 · HCl, purum ≥

98%), Goat Serum were from Sigma-Aldrich (Oakville, ON). All buffer solutions were prepared

using a deionized water purification system (MilliQ, 18 MΩ.cm) and were autoclaved prior to

use. The buffer solutions included 100 mM tris-borate buffer (TB, pH 7.4) and 100 mM

bicarbonate buffer (BIC, pH 8.3).

Synthesis of OD539 (3-(4-Carboxybutyl)-2-[4(N,N-dimethylamino)styryl]benzoxazolium).

The synthesis of this dye has been reported elsewhere10. Briefly, 2-methylbenzoxazole (3.00 ml,

25.0 mmol) and 5-bromovaleric acid (9.00 mg, 50 mmol) were suspended in 10 mL 1,2-

dichlorobenzene. The resulting mixture was stirred and heated at 120 oC overnight. The reaction

mixture was cooled to room temperature and 200 mL of dichloromethane was added. The

resulting suspension was stirred at room temperature for 2 hours. The precipitate was then

filtered, washed and dried resulting in white powder. The white powder (312 mg, 1.00 mmol)

was then mixed with 4-dimethylaminobenzaldehyde (150 mg, 1.00 mmol) and suspended in 10

mL acetic anhydride. The suspension was heated and stirred at 120 oC for 30 min. At the end of

6

the 30 min, 10 mL of purified water (MilliQ, 18.2 MΩ.cm) was added and the resulting solution

was heated for another 30 min. The solvent was then evaporated at reduced pressure and 100 mL

acetone was added to the residue. The resulting suspension was allowed to stand at room

temperature for 30 min. The precipitate was filtered, washed with acetone and dried under

reduced pressure giving a reddish brown powder. The mass of the synthesized dye was found to

be 365.06 ([M-Br]+) using ESI MS.

The resulting dye (8.9 mg, 20 μmol) was mixed with N-Hydroxysuccinimide (4.6 mg 40 μmol )

and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride (7.7 mg, 40 μmol) in 0.5 mL

dimethylformamide and stirred at room temperature overnight. The resulting clear solution was

stored in a freezer.

Attaching dye to the DNA probe

DNA probe was from Integrated DNA Technologies (IDT) and was provided with two internal

amine linkers, AmC6dT or UniLinkTM, and a biotin attached to the 5’end. AmC6dT is a modified

nucleotide with an amine attached to the nucleobase by a 6 carbon linker. UniLinkTM has an

amine that is attached to the phosphate backbone by a 6 carbon linker. The dye solution was

added to the diamino modified DNA probe at 50 mole equivalents in 100 mM bicarbonate

buffer at pH 8.3. The resulting solution was shaken using a vortex mixer and left overnight (at 4

oC). The solution was then centrifuged at 8000 rpm for 3 min. The supernatant was collected and

purified using a NAP5 column. The concentration of the DNA probe was determined using the

DNA absorption band at 260 nm (Molar Extinction Coefficients were provided by IDT).

Preparation of QD-probe conjugates

7

The doubly labeled DNA probes were mixed with streptavidin coated green QDs at 70

equivalents in 100 mM Tris Borate (TB) buffer with 20 mM NaCl at pH 7.4. These QDs

accommodate on average 46 probes per QD [13]. The solution was mixed on a shaker platform

for 45 min. 100 kD spin filters were used to remove excess DNA probes in solution. The

immobilization of the DNA probes was confirmed using Gel Electrophoresis (Supporting

Information).

Positioning of Dyes on Oligonucleotide Probes

Two types of probes were ordered from IDT; one contained two thymine bases each with amino

modified linkers (Amino Modifier C6 dT) and the other was modified with two Uni-LinkTM

Amino Modifiers. The modified thymine bases were adjacent, located either in the middle of the

probe sequence or one quarter of the distance from the terminus of the oligonucleotide sequence.

The points for dye attachment were either immediately next to each other or separated by one

nucleotide. The dye-labeled probes were conjugated to the QDs and the resultant QD-probe

conjugates were incubated with fully complementary (FC) target sequences at 2 μM

concentration for 2 hours in a TB buffer with 100 mM NaCl. A summary of the probe sequences

is shown in Table 1.

Table 1. SMN1 oligonucleotide sequences used in the assayName Sequence50-1 Probe 5’ Biotin-TTG ATT T/Amine/G /Amine/CT GAA ACC C 3’

50-1 FC tgt 5’ TAC TGG CTA TTA TAT GGG TTT CAG ACA AAA TCA A 3’

50-1 1BPM-0 tgt 5’ TAC TGG CTA TTA TAT GGG TTT CAC ACA AAA TCA A 3’

50-1 1BPM-1 tgt 5’ TAC TGG CTA TTA TAT GGG TTT CTG ACA AAA TCA A 3’

50-1 1BPM-2 tgt 5’ TAC TGG CTA TTA TAT GGG TTT AAG ACA AAA TCA A 3’

8

50-1 1BPM-3 tgt 5’ TAC TGG CTA TTA TAT GGG TTT CAG ACA AAA GCA A 3’

25-1 Probe 5’ Biotin -ATT T/Amine/G /Amine/CT GAA ACC CTG T 3’

25-0 Probe 5’ Biotin -ATT /Amine//Amine/GCT GAA ACC CTG T 3’

25-1 FC tgt 5’ TCC TTT ATT TTC CTT ACA GGG TTT CAG ACA AAA T 3’

tgt = target, FC = fully-complementary, 1 BPM = 1 Base Pair Mismatch, NC = noncomplementary, /Amine/ = modified thymine (AmC6dT) or UniLinkTM. The mismatched base in 1 BPM tgt is bolded, italicized and underlined.

Sensitivity and Selectivity assays

The 34 and 90 base length target oligonucleotides were from Integrated DNA Technologies

(Table 1). Calibration curves were constructed using solutions of 0.01 to 2 μM target

concentrations. The targets were incubated with the QD-probe conjugates in 100 mM TB buffer

with 100 mM NaCl at pH 7.4 for 1-2 hours. The quantum dots were excited at 405 nm using a

fluorescence spectrometer and photoluminescence (PL) was measured between 500 - 600 nm.

The selectivity of the assay was evaluated by comparing the PL for fully complementary (FC)

target with one containing a base pair mismatch (1BPM). The target with 1BPM was incubated

with the QD-probe conjugate in a 100 mM TB buffer at pH 7.4 with 100 mM NaCl for 2 hours.

The effect of the location of the 1BPM was investigated by placing the dye and mismatch at

various distances with up to 3 nucleotides of separation. This experiment was repeated in the

presence of 75% v/v goat serum.

9

Figure 2. a) Normalized absorption and PL of the dye labeled probe measured in the absence and the presence of the FC target. The absorption (dashed line) and PL (solid line) are indicated by red and blue for the nonhybridized probe and the hybrid, respectively. b) The excited state of the dimer splits into two energy levels due to the interaction of the transition dipoles. The out-of-phase alignment of the dipoles results in the lower energy level, whereas in-phase alignment of the dipoles results in the higher energy level. The transition from the ground state to the lower energy level is forbidden. After excitation to the allowed upper level, rapid internal conversion takes place to the lower level, which is followed by non-radiative relaxation to the ground state15.

RESULTS AND DISCUSSION

The absorption and emission spectra of the doubly labeled probes were measured in the presence

and the absence of the fully complementary target. Two absorption bands were present in both

cases in a range of 400-550 nm (Figure 2a). The band centered with the shorter wavelength (482

nm) was more prominent when the probes were in the single-stranded state. The longer

wavelength absorption band (520 nm) dominated when the fully complementary target

hybridized to the probe. This latter absorption band was also prominent in the spectrum of a

probe labeled with only a single dye molecule (Supporting Information). The difference in the

photophysical properties of the dyes in the presence and the absence of the target are due to the

formation of a dimeric structure between the dyes9. The self-association of the dye molecules in

10

aqueous solutions occurs due to π stacking interactions between the conjugated systems of the

dyes, which minimizes exposure to water14. The shift towards a shorter wavelength upon

formation of a dimeric structure is a characteristic of H-aggregates. The formation of the H-

aggregate allows for excitonic interaction between the dye molecules, which in turn suppresses

fluorescent emission10 (Figure 2b). Exciton coupling theory suggests that the dye molecules are

treated as point dipoles and the excited state splits into two energy levels due to the interaction of

transition dipoles15. The higher energy level results from the in-phase alignment of the two

dipoles, whereas the out-of-phase alignment gives rise to the lower energy level15. Absorption of

an incoming photon results in the transition to the upper excitonic state followed by rapid

deactivation to the non-emissive lower excitonic state15. Hybridization of the fully

complementary target results in the dissociation of the dimer, leading to the disruption of the

excitonic interaction and restoration of the fluorescence emission. Fluorescence intensity is

directly affected by the quantum yield, and the quantum yield is expressed by the ratio of the

sum of the radiative relaxation over the sum of the non-radiative relaxation rate constants11. The

disruption of the excitonic interaction leads to increase in the quantum yield and the subsequent

increase in fluorescence intensity. The quantum yield of the probe-target hybrid was 1.6 times

the quantum yield of the probe in the single-stranded state which is lower than expected for a

derivative of a thiazole orange dye in a similar environment9. The relatively small change in the

quantum yield may be a consequence of incomplete dissociation of the dimer upon hybridization

of the target. The presence of two emissive populations, dimer and monomer, was evident in

lifetime measurements (Supporting Information). The fluorescence decay profile of the probe in

the double stranded state was fit well with a double exponential function; whereas, the probe in

11

the single stranded state could be fit to a single exponential function. The presence of the dimeric

form is also supported by the absorption spectra. The shorter wavelength absorption band is

present as a shoulder in the spectrum of the double stranded state and absent in the spectrum of

the probe labeled with a single dye (Supporting Information).

The PL associated with the hybridization of the target to the probe was quantified using a

ratiometric approach based on a modified FRET ratio (MFR) (Supporting Information). The

12

Figure 3. Two types of linkages were used to attach the dye to the DNA probe, amine modifier UniLinkTM (a) and an amino modified thymine base (b) The UniLinkTM conjugates the dye to the phosphate backbone of the oligonucleotide; whereas, the modified thyme conjugates the dye to the nucleobase. The MFR for the probe containing the modified thymine bases was 5.2 times the MFR of the probe labeled with the UniLinkTM. (c-d) Absorption spectra of the doubly labeled probe in the presence and the absence of the target, as well as of a singly labeled probe for “50-uni” and “50-1”, respectively.

change in the magnitude of the MFR provided an indication of the effect of the type of linkage

used to attach the dye to the probe, the location of the dye within the probe sequence, as well as

the distance between the dyes in the presence and the absence of the target.

The type of linkage used to attach the dye to the probe had a significant effect on the PL of the

dye. There were two types of commercially available linkages used; both contained a six carbon

aliphatic spacer arm. A Uni-LinkTM amino-modifier provided for attachment of an amine to the

phosphate backbone of an oligonucleotide sequence (Figure 3). A modified thymine nucleobase,

AmC6dT, provided for conjugation to the nitrogenous base. The MFR of the probe containing

the AmC6dT (50-1 probe) was 4.6 times higher than the MFR of the probe containing the Uni-

LinkTM (50-uni probe) (Figure 3a). The lower PL of the UniLinkTM probes could result from steric

restrictions imposed on the dyes. Unlike the dyes directly attached to the nucleobases, the dyes

attached to the DNA via the UniLinkTM spacer might not have sufficient range of motion, hence

preventing the dyes from intercalating into the DNA duplex. This is evident in the absorption

spectrum of the “50-uni” probes as the characteristic long wavelength absorption peak of the

monomeric state emerged as only a shoulder after the hybridization of the fully complimentary

target, and the short wavelength absorption band that is characteristic of the dimeric state

remained prominent (Figure 3 c-d).

13

Figure 4. The signal from the dye in the presence of the FC target was not significantly impacted due to variation of the distance between the locus of the dyes and the QD. In one configuration the dye was attached to the center of the probe (50-AmC6dT, a), and in another configuration was attached a quarter of the distance along the probe length (25-AmC6dT, b). The dyes were attached using the amine modified thymine, C6dT. The ratio between the MFR of the 50-AmC6dT probe and the 25-AmC6dT probe was 1:1.

The location of the dye within the probe sequence was chosen relative to the 5’ end of the

sequence (Figure 4) so that FRET efficiency would be maintained. In the case of the probes

referred to as “25-1”, the dyes were attached to the 5th and the 7th nucleotides, and this represents

one quarter of the length of the sequence. The probes referred to as the “50-1” located the

conjugated dyes in the middle of the sequence. Based on Equation 1, the location of the dye with

respect to the QD was expected to have a significant effect on the signal intensity. However,

there were no significant differences between the MFR for the two dye configurations. In order

to rule out the possibility of the probes collapsing on the surface of the QDs, unlabeled probes of

the same sequence were hybridized with targets labeled with Cy3 dye at the 3’and the 5’ end.

The FRET ratio observed for the proximal target was 6 times the FRET ratio of the distal target,

which suggests that the probes do not fold onto the nanoparticle surface. The results shown in

14

Figure 4 suggest that any increase in the FRET efficiency due to a closer proximity of the dye to

the donor was largely offset by weakened intercalation of the dyes due to being close to the

terminus of the hybrid sequence16. The signal intensity of the dye is expected to increase upon

hybridization of the target as it causes disassociation of the non-emissive dimer; however, it was

previously noted that partial intercalation of the dye caused only a 50% increase in the signal

intensity16.

Figure 5. (a-b) The effect of the distance between two dyes was investigated by attaching the dyes to nucleotides that were immediately adjacent (25-0) or that were separated by one nucleotide (25-1). The dyes were attached to the probe one quarter of the way along the length of the sequence using the amine modified thymine AmC6dT. (c-d) Absorption spectra for the probe in the single stranded state, double stranded state and a single dye molecule for “25-1” and “25-0”.

15

The distance between the dyes had a significant effect on the signal intensity. The MFR of the

probes labeled with dyes that were separated by one nucleotide (25-1 probes) was a factor of 4.5

greater than the MFR of the probes labeled with dyes that were immediately adjacent (25-0

probes) (Figure 5 a-b). It was anticipated that increasing the distance between the dyes would

weaken the excitonic interaction between the dyes, and this was supported by the absorption

spectra. In case of the “25-0” probe, the short wavelength absorption band did not reduce in

intensity and the longer wavelength absorption band emerged only as a shoulder after the

addition of the fully complementary target, which indicates that the excitonic interaction was not

sufficiently disrupted (Figure 5 c-d).

Figure 6. a) Concentration-response showing MFR response to 34mer and 90mer FC target. b) Fluorescence emission spectrum corresponding to a - QD, b - QD-probe conjugate, c-k increasing concentrations of target.

The quantitative response of the assay was investigated using FC targets of 34mer and a 90mer

length. In case of the 34mer target, the response increased linearly with target concentration

ranging from 0.01 μM (2 pmol) to 0.75 μM (150 pmol), corresponding to a dynamic range of one

order of magnitude (Figure 6). The limit of detection (LOD) of the assay was determined

experimentally to be 2 pmol. In case of the 90mer target, the dynamic range, the detection limit

16

and magnitude of the signal was similar to that of the 34mer target (Figure 6). The similarity of

hybridization efficiency of the shorter and longer targets is attributed to the angle of deflection

between the strands provided by the radius of curvature of the nanoparticle interface17. The

increase in distance between the strands on a surface of high curvature reduces steric

interactions, thus facilitating hybridization efficiency17. The quantitative response of the assay

was also investigated in a complex matrix consisting of 75% v/v Goat Serum (Supporting

Information). Despite some reduction of signal intensity, the dynamic range and the limit of

detection were found to be similar to those obtained in buffer solution.

The primary intention of this work was to integrate intercalating dyes into single-stranded

oligonucleotide probes in a manner that could be used for transduction of hybridization by a QD-

FRET mechanism. A further opportunity involving improvement of detection level was

examined. Excitonic interaction between the dye molecules in the doubly labeled probes caused

quenching of the fluorescence, thus reducing background. The resultant increase in the difference

between the signals obtained from the hybridized and the non-hybridized states allowed for

improvement of the limit of detection. There was higher background in the case of the singly

labeled probes, and the LOD was improved by 5 with doubly labeled probes (Supporting

Information). In addition, the sensitivity (slope of response) using the doubly labeled probes was

higher than observed for the singly labeled probes (Supporting Information). The sensitivity of

the configuration containing QDs and doubly labeled probes was also compared to the sensitivity

of a system in the absence of the QDs where the doubly labeled probes were excited directly by

the excitation source (Supporting Information). Results showed a 34% increase in the sensitivity

of the system for probes that were excited using FRET from QD donors, with equivalent LOD,

indicating that the advantages of QD-FRET excitation were achieved without sacrificing

17

analytical performance. In addition, the precision in the measurements were improved in the case

of direct excitation of the doubly labeled probes.

Figure 7. Selectivity of the assay was evaluated by comparing the MFR of FC and a 1BPM target. (a) The mismatch was located next to the labeled nucleotide (Table 1). (b) The absorption spectrum of the 1BPM does not have the characteristic peak of a probe labeled with a single dye molecule at 520 nm. (c) With increasing distance between the mismatch and the dye the signal intensity and the MFR increased.

Single nucleotide polymorphism (SNP) discrimination was used as a proxy to evaluate the

selectivity of the assay. A sequence was chosen for SNP discrimination in which the 1BPM was

located next to the amine modified thymine that was linked to the dye (1BPM-0). The MFR for

this probe was -0.08±0.03, indicating that the analytical signal was lower than the background

18

signal that was observed in the absence of the target (Figure 7a). The magnitude of the MFR was

observed to behave as expected as the distance between the 1BPM was moved by 1 or 2

nucleotides (Figure 7c) away from the amine modified thymine conjugation sites. The formation

of the mismatched base pair in the 1BPM-0 probe results in the decrease in the fluorescence

intensity because the mismatched nucleotide located next to labeled nucleotide serves as a

binding site for the dye9. The mismatched base pair causes the dissociation of the dyes from the

DNA structure by lowering the binding affinity to the DNA. The dissociated dyes form a non-

emissive bichromophoric aggregate which in turn results in the decrease of fluorescence intensity

9, as is evident from the absorption spectrum of the dye in the presence of the 1BPM target where

the characteristic monomeric absorption peak at 520 nm is missing (Figure 7b). The experiment

was repeated in the presence of goat serum. The results showed reduction in signal due to

increase in the background; however, the same trend was observed (Supporting Information,

Figure S3).

CONCLUSIONS

An intrinsically-responsive FRET transduction method to detect hybridization was investigated

using dye labeled oligonucleotide probes that were immobilized on the surface of QDs. Two

commercially available linkers were assessed for concurrent conjugation of two identical

thiazole orange derivatives at various locations on single-stranded probe oligonucleotide. Amino

modified thymine provided for better signal-to-noise in comparison to attachment of dyes to the

phosphate backbone of oligonucleotide probes. It was shown that attachment of the dyes closer

to the QDs did not result in the improvement of the signal for the relatively short oligonucleotide

probes that were used. This was attributed to the possibility that the increase in the FRET

19

efficiency due to decrease in the distance between the donor and acceptor was counteracted by

the decrease in fluorescence intensity due to partial intercalation of the dyes as they were located

near the end of the sequence where the hybrid was labile. The impact of the distance between

the dyes on signal generation was explored. The dyes were attached either next to each other or

separated by one nucleotide. The latter provided for improved signal intensity, as increasing the

distance between the dyes further weakened the excitonic interaction so that they could more

effectively intercalate into the DNA duplex. The system was also assessed for quantitative

response to increasing concentrations of a fully complementary 34mer target and a 90mer target,

indicating pmol detection levels with similar response to both sequences even though they

differed in length. The selectivity of the assay was investigated using a 1BPM target, establishing

that the location of the mismatch was important because the mismatch could serve as an

interaction site for adjacent dyes.

ACKNOWLEDGEMENTS

The authors gratefully acknowledge the Natural Sciences and Engineering Research Council of

Canada (NSERC) for financial support of this research. The authors would also like to thank

Samantha Tabone for technical assistance with some of the research work and for helpful

discussion.

ASSOCIATED CONTENT

Supporting Information

Description of instrumentation used, characterization of the FRET pair and equations used in the

data analysis, gel electrophoresis results, additional absorption spectrum and SNP discrimination

in 75% v/v goat serum solution.

20

REFERENCES

1. Hilal, H.; Taylor, J. A. J. Biochem. Biophys. Methods. 2008, 70, 1104–1108

2. Sang, F.; Ren, J. J. Sep. Sci. 2006, 29, 1275 – 1280

3. Bahrami, A. R.; Dickman, M. J.; Matin, M.M.; John R. Ashby, J. R.; Brown, P. E.; Conroy, M.J.; Fowler,G.J.S.; Rose,J.P.; Sheikh, Q.I.; Yeung, A. T.; Hornby D.P. Anal. Biochem. 2002, 309, 248–252

4. Bengtsson, M.; Karlsson, H. J.; Westman, G.; Kubista, M. Nucl. Acids Res. 2003, 31, e45

5. Chi, C.; Yeh-Hsing Lao, Y.; Li, Y.; Chen, L. Biosens. Bioelectron. 2011, 26, 3346–3352

6. Zhang, H.; Zhou, D.; Chem. Comm. 2012, 48, 5097–5099

7. Kaloyanova, S.; Trusova,V. M.; Gorbenko, G.P.; Deligeorgiev, T. J. Photochem. Photobiol. A: Chem. 2011, 217,147–156

8. Armitage, B. A. Top Curr. Chem. 2005, 253, 55–76

9. Ikeda, S.; Kubota, T.; Kino, K.; Okamoto, A. Bioconjugate Chem. 2008, 19, 1719-1725

10. Ikeda, S.; Kubota, T.; Yuki, M.; Okamoto, A. Angew. Chem. Int. Ed. 2009, 48, 6480 –6484.

11. Skoog, D.A.; West D. M.; Holler F.J.; Crouch S.R. Fundamentals of Analytical Chemistry.

8th ed. Brooks/Cole: 2001, pg 707-875.

12. Noor, O. M.; Shahmuradyan, A.; Krull, U. J. Anal. Chem. 2013, 85, 1860−1867.

13. Noor, O.M.; Tavares, A. J.; Krull, U.J. Anal. Chim. Acta. 2013, 788, 148–157

14. Armitage, B. A. Top Curr. Chem. 2005, 253, 55-76.

15. Kasha, M.; Rawls, R.; El-Bayoumi, M. A. Pure and Appl. Chem. 1965, 11, 371-392.

16. Wanga, D. O.; Okamoto, A. J. Photochem. Photobio. C: Photochem. Rev. 2012, 13, 112– 123

21

17. Noor, O. M.; Hrovat, D.; Moazami-Goudarzi, M.; Espie, G., S.; Anal. Chim. Acta. 2015, 885, 156–165

22