tspace.library.utoronto.ca  · Web viewTo explore whether the rate of hybridization was limited by the diffusion of the QDs, kinetic experiments for the detection of 0.5 µM of HPRT1

A paper-based resonance energy transfer nucleic acid hybridization assay using upconversion

nanoparticles as donors and quantum dots as acceptors

Samer Doughan, Uvaraj Uddayasankar 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: [email protected]

Abstract:

Monodisperse aqueous upconverting nanoparticles (UCNPs) were covalently immobilized on aldehyde

modified cellulose paper via reduction amination to develop a luminescence resonance energy transfer

(LRET)-based nucleic acid hybridization assay. This first account of covalent immobilization of UCNPs on

paper for a bioassay reports an optically responsive method that is sensitive, reproducible and robust.

The immobilized UCNPs were decorated with oligonucleotide probes to capture HPRT1 housekeeping

gene fragments, which in turn brought reporter conjugated quantum dots (QDs) in close proximity to

the UCNPs for LRET. This sandwich assay could detect unlabeled oligonucleotide target, and had a limit

of detection of 13 fmol and a dynamic range spanning nearly 3 orders of magnitude. The use of QDs,

which are excellent LRET acceptors, demonstrated improved sensitivity, limit of detection, dynamic

range and selectivity compared to similar assays that have used molecular fluorophores as acceptors.

The selectivity of the assay was attributed to the decoration of the QDs with polyethylene glycol to

eliminate non-specific adsorption. The kinetics of hybridization were determined to be diffusion limited

and full signal development occurred within 3 minutes.

Keywords: Upconversion Nanoparticles, Quantum Dots, Luminescence Resonance Energy Transfer,

Nucleic Acid Hybridization, Bioassay, Paper

1

1. Introduction:

Lanthanide doped upconverting nanoparticles (UCNPs) have attracted much attention for

applications in bioanalysis due to their unique optical properties. Upconversion is based on the

sequential absorption of two or more photons in the IR region of the electromagnetic spectrum

followed by the narrow band emission of radiation of higher energy within the UV to NIR region. IR

excitation minimizes autofluorescence from biological material, and reduces optical background that is

commonly associated with an excitation source that operates in the UV or visible wavelength range. In

addition, lanthanide doped UCNPs emit multiple narrow and well-defined emission peaks suitable for

multiplexed optical analysis[1]. UCNPs have been used in bioassays as passive labels, and as energy

donors in LRET for the detection of nucleic acids[2-4], proteins[5-7] and metal ions[8-10]. While UCNP

LRET-based assays offer access to a ratiometric approach that provides for good precision, they suffer

from LRET efficiencies that are generally well below 50%. Improving the LRET efficiency provides higher

sensitivity and lower detection limits of bioassays. Reported strategies to improve LRET efficiency

include surface decoration of UCNPs[11], adjustment of donor and acceptor distance[12] and

optimization of LRET acceptor properties[13]. We have previously demonstrated the use of a sandwich-

based assay format for the detection of thrombin where a dense monolayer of UCNPs deposited onto a

glass surface allows a single LRET acceptor to interact with multiple donors[5]. The multiple donor -

acceptor interactions at the surface provided about 4 fold enhancement of the LRET ratio.

To further improve sensitivity and dynamic range, in this work UCNPs were immobilized on paper to

make use of the large surface area associated with the three dimensional matrix. Paper based assays

have attracted much attention due to their low cost, fluid transport via capillary action, and easy

modification and patterning[14]. More importantly, the three dimensional nature of paper was reported

by Noor et al. to be capable of providing more than a 10 fold enhancement in fluorescence resonance

energy transfer (FRET) ratio for immobilized quantum dots (QDs) and dye acceptors in a label free

2

nucleic acid hybridization assay[15, 16]. The enhancement was attributed to the large available surface

area, in combination with the contraction of wet paper upon drying that brought neighboring donors

and acceptors into closer proximity[16]. Zhou et al. adsorbed UCNPs on paper for the detection of dye

labeled oligonucleotides, however, the assay was limited in sensitivity and selectivity[17, 18]. Herein, we

report a novel design for the sensitive and selective detection of unlabeled target oligonucleotide on

paper using covalently immobilized UCNPs as donors and QDs as acceptors. QDs are more photostable,

offer higher extinction coefficients and wider absorption spectra than molecular dyes, and are known as

excellent LRET acceptors[19].However, QDs have broad spectral absorption profiles and this has limited

their use as acceptors. This is primarily because the wavelengths that are used to excite donors will

often concurrently directly excite QDs, making it impossible to excite the QDs only by resonance energy

transfer from the donor[19]. Use of QDs as acceptors is typically limited to chemiluminescence energy

transfer (CRET)[20] and bioluminescence energy transfer (BRET)[21] where no excitation source is used.

QDs have also been used as acceptors with lanthanide complexes as donors in time gated

measurements[22]. Herein, QDs are effectively used as LRET acceptors without the need for time gated

measurements. An epifluorescence microscope equipped with a continuous 980 nm laser provides for

photoluminescence from UCNPs that in turn can excite QDs, where the intensity of acceptor emission is

measured using a photomultiplier tube in conjunction with appropriate band pass filters. The narrow

and well defined emission peaks of both donor and acceptor makes it possible to collect LRET sensitized

QD photoluminescence in the absence of any donor background.

This work presents the first account of use of covalently immobilized UCNPs on paper as LRET donors for

the optical detection of unlabeled nucleic acid targets (Figure 1). Oligonucleotide probes decorating the

UCNPs capture HPRT1 housekeeping gene fragments. An unhybridized segment of the HPRT1 target in

turn hybridizes with an oligonucleotide reporter that is conjugated to QDs. This results in localization of

QDs in close proximity to the UCNPs for LRET. The kinetics of hybridization are optimized, and non-

3

specific adsorption by QDs is eliminated to build a hybridization assay that offers speed and high

selectivity[17].

Figure 1: The strategy for the detection of target DNA. Probe oligonucleotide is conjugated to an UCNP, and the UCNP is immobilized on a paper substrate. Target oligonucleotide serves to bridge probe oligonucleotide on the UCNP and reporter oligonucleotide on the QD. Excitation of the UCNP at 980 nm provides for luminescence that excites the QD. The paper substrate is prepared to localize 3 reaction spots that are defined by wax printing. The orange, black and green nucleic acid strands represent the probe, target and reporter strands, respectively.

2. Experimental

A full list of Materials and Instrumentation can be found in the supporting information.

2.1 Synthesis of NaYF4: 0.5% Tm3+, 30% Yb3+/NaYF4 core/shell UCNPs

Core NaYF4: 0.5% Tm3+, 30% Yb3+ UCNPs were synthesized according to previous reports[23]. In short,

0.4562, 0.2534, and 0.0042 g of Y(CH3CO2)3.xH2O, Yb(CH3CO2)3.4H2O and Tm(CH3CO2)3.xH2O were stirred

4

O

O

N

N

O

O

N

OH

HO

NOP O-

OO-

-O

O

-O

PO

O PO-

O O-

-O

O

O-PO

575 nm

O

O

N

N

O

O

N

OH

HO

N

OP O-OO-

-O

O

-O

PO

O PO-

O O-

-O

O

O-

PO

OO

NNO

ON OH

OH

N

OP

O-

O

-O

-O

O

O-

PO

OP

O-O

-O

-O

OO-

PO

980 nm

UCNP

Q

D

gently in 30 mL octadecene and 12 mL oleic acid (OA) under vacuum at 115 °C for 30 min. The mixture

was then cooled to 50 °C under argon before a 20 mL methanol solution containing 0.20 g NaOH and

0.30 g NH4F was added. The mixture was stirred for 30 min and was then heated to 75 °C to evaporate

the methanol. The temperature was rapidly increased to 300 °C and maintained for one hour. The

solution was allowed to cool to room temperature before the core UCNPs were precipitated using

ethanol and centrifugation at 4500 rpm. The UCNPs were resuspended in hexanes and recaptured with

ethanol and centrifugation. Core UCNPs were stored in hexanes at 4 °C.

The core UCNPs were subsequently capped with a NaYF4 shell. In a three neck round bottom flask,

0.5738 g of Y(CH3CO2)3.xH2O were stirred gently in 30 mL octadecene and 12 mL OA under vacuum at

115 °C for 30 min. The temperature was lowered to 80 °C with the mixture under argon, and core UCNPs

in hexanes were added. The reaction temperature was maintained until the hexane was evaporated

after which the reaction was cooled to 50 °C. A volume of 20 mL methanol solution containing 0.14 g

NaOH and 0.26 g NH4F was added and stirred for 30 min. The temperature was increased to 75 °C to

evaporate the methanol, and then the temperature was rapidly increased to 300 °C and maintained for

one hour. Core/shell UCNPs were precipitated using ethanol and centrifugation at 4500 rpm. The UCNPs

were resuspended in hexanes and recaptured with ethanol and centrifugation three times. The

core/shell UCNPs were stored in hexanes at 4 °C.

2.2 Preparation of water soluble UCNPs

Oleic acid capped core/shell UCNPs were made water soluble by ligand exchange with o-

phosphorylethanolamine (PEA). In a typical reaction, 100 mg of OA-UCNPs in 2 mL hexanes were mixed

with 400 mg of PEA and 1 mL of tetramethylammonium hydroxide (TMAH) in 10 mL of ethanol [5]. The

reaction was allowed to stir vigorously overnight in a capped glass vial at 70 °C. The PEA coated UCNPs

were recovered by centrifugation at 3500 rpm. The UCNPs were washed three times by sonication in

5

ethanol for 5 min before the addition of hexanes and centrifugation at 4500 rpm. The washed PEA-

UCNPs were suspended in 10 mL of water before passing through a 0.2 µm poly(ether sulfone) (PES)

syringe filter to remove any aggregates. The aqueous UCNPs were stored at 4 °C in excess PEA.

2.3 Modification of paper and immobilization of PEA-UCNPs

Reaction zones in the form of spots with an inner diameter of 0.3 cm were defined on paper by wax

printing. Hydroxyl groups of cellulose in the defined spots were oxidized using an aqueous solution

containing 30 mg mL-1 lithium chloride and 10 mg mL-1 sodium periodate. Each spot was treated with 10

µL of the oxidizing solution and was allowed to incubate at 50 °C until dryness. The papers were washed

with purified water in a 50 mL tube for 5 min and were dried in a desiccator.

PEA-UCNPs were immobilized on aldehyde functionalized paper via reduction amination. Excess PEA

from the stock PEA-UCNP solution was first removed using a 100 kDa centrifugal filter. A 1.5 mg mL -1

PEA-UCNP solution was then prepared in HEPES buffer (100mM, pH 7.2) containing 1 mM sodium

cyanoborohydride. Into each spot, 5 µL of the UCNP solution was pipetted and the spots were allowed

to incubate for 10 minutes before they were washed with a 0.1% v/v aqueous Tween ® 20 solution for 5

minutes. The papers were then washed with purified water for 2 minutes before they were dried in a

desiccator.

2.4 Preparation of hexahistidine functionalized oligonucleotides and mPEG thiol

Modifications were based on previous reports[24, 25]. The thiol modified reporter nucleic acids (Table

1), diluted in 1 x PBS, were incubated with 500 molar equivalents of dithiothreitol (DTT) for 1 hour at

room temperature to reduce the disulfides into sulfhydryl moieties. The nucleic acids were then isolated

from excess DTT by extracting the aqueous solution with ethyl acetate four times. To functionalize the

nucleic acid with the peptide, 10 molar equivalents of the maleimide-functionalized peptide (dissolved

6

in dimethyl sulfoxide; (6-Maleimidohexanoic acid) – G(Aib)GHHHHHH) was added, and the solution was

allowed to shake for 12 hours at room temperature. Excess peptide was then removed by passing the

nucleic acid solution through a NAP-5 desalting column, as per the manufacturer’s instructions. The

purified nucleic acid was quantified using UV-vis spectroscopy and then stored at -20 0C.

The thiol functionalized poly(ethylene glycol) (PEG) methyl ether (Mn = 6000 g mol-1) was incubated with

10 molar equivalents of tris (2-carboxyethyl) phosphine (TCEP) for 2 hours to reduce any disulfides.

Excess TCEP was removed using a NAP-5 desalting column, and the thiol mPEG was then incubated with

10 equivalents of maleimide-functionalized peptide for 12 hours. Un-reacted peptide was subsequently

removed using a NAP-5 desalting column.

2.5 Preparation of QD-reporter conjugates

Alkyl 575nm CdSxSe1-x/ZnS core/shell QDs were obtained from Cytodiagnostics Inc. (Burlington, ON,

Canada) and were made water soluble via ligand exchange with reduced L-glutathione (GSH)[15]. In one

glass vial, 75 µL of 10 µM alkyl QDs was suspended in 2mL chloroform, and in another glass vial, 0.2 g of

GSH was dissolved in 600 µL of TMAH. The solution containing QDs was then added drop-wise to the

GSH solution while swirling. The mixture was allowed to sit overnight and the QDs were collected by

centrifugation. The GSH QDs were washed three times by resuspension in pH 9.25 borate buffer (50

mM, 100 mM NaCl) and collection with ethanol and centrifugation. The QDs were then stored in pH 9.25

borate buffer (50 mM, 100 mM NaCl) at 4 °C.

The hexahistidine modified oligonucleotides were incubated with GSH QDs at the desired QD:DNA ratio

for one hour in pH 9.25 borate buffer (50 mM, 100 mM NaCl). The solution was then treated with

hexahistidine modified PEG at 15x excess the amount of QDs for one hour. The solution was washed

three times using a 100 kDa centrifugal filter. The PEG stabilized DNA conjugated QDs were collected

and stored in pH 9.25 borate buffer (50 mM, 100 mM NaCl) at 4 °C.

7

2.6 Quantification of the average number of oligonucleotides per QD

After the QDs were incubated with the desired amount of hexahistidine modified DNA, the average

number of oligonucleotides conjugated per QD was determined using a well-established method[20].

The QD conjugates were purified as previously described and their absorbance spectra was obtained.

The contribution of the oligonucleotides to the absorbance at 260nm was obtained by subtracting the

contribution provided by the presence of the QDs. The latter was obtained from a normalized

absorbance spectrum of the QDs in the absence of DNA. The average number of oligonucleotides per

QD was then determined by the ratio of the concentrations of oligonucleotides to QDs.

2.7 Hybridization and detection on paper substrates

Spots containing immobilized PEA-UCNPs were treated with 3 µL aliquots of a 2 mM aqueous solution of

NHS-PEG4-biotin. The spots were allowed to dry, and the paper was washed for 2 min in purified water

before it was dried in a desiccator. Subsequently, 5 µL aliquots of a 20 µM avidin solution in HEPES

buffer (100 mM, pH 7.2) was pipetted onto the spots. The paper was allowed to dry before it was

washed for 2 minutes with borate buffer (50 mM, pH 9.25) and dried again. Each spot was then treated

with a 5 µL of solution containing 10 µM biotinylated oligonucleotide probe and 10 µM biotinylated

oligonucleotide of non-complementary sequence but of the same length to ensure dispersion of the

probe on the surface. The surface of the paper was then blocked by treatment with a 20 µM solution of

unlabelled oligonucleotide to minimize non-specific adsorption and the paper was washed for 2 minutes

in pH 9.25 borate buffer (50 mM, 100 mM NaCl). After the paper was dried, 3 µL solutions with the

desired DNA target concentration ranging from 5 nM to 2 µM were introduced. The paper was allowed

to dry before the addition of a 3 µL solution containing reporters consisting of PEG stabilized DNA

conjugated to GSH-QDs. The solution was allowed to incubate for 5 minutes and was washed with pH

8

9.25 borate buffer (50 mM, 100 mM NaCl) containing 0.1% v/v Tween for 2 minutes. The paper was

allowed to dry in a desiccator before it was imaged using an epi-fluorescence microscope.

Table 1: Oligonucleotide Sequences used in the Hybridization Assays

Name Sequence

HPRT1 probe 5'-Biotin- CAAAATAAATCAAGGTCA-3'

HPRT1 reporter 5'-TAACCTGGTTCATCATC- Thiol-3'

HPRT1 target 5'-GATGATGAACCAGGTTATGACCTTGATTTATTTTG-3'

HPRT1 1BPM target 5'-GATGATGAACCAGGTTATGACCTTGTTTTATTTTG-3'

Probe Spacer 5'-Biotin-CAGACACACAGGGTGAAT-3'

Non-complementary DNA 5'- ACACACACACACACACACACACACA-3'

3. Results and Discussion:

3.1 UCNP Synthesis and Immobilization:

Core-shell OA-UCNPs were synthesized using the seeded growth method and were made water soluble

via ligand exchange with PEA at an elevated temperature. The use of TMAH allowed for the

deprotonation of the phosphate groups of PEA promoting their coordination to the surfaces of the

UCNPs. Phosphate groups have been shown to coordinate strongly to the surface of UCNPs and are

frequently used for ligand exchange [5, 26]. The TEM image in Figure S1 shows aqueous monodisperse

UCNPs that were predominantly cubic in shape. The average diameter was 24.2 ± 3.4 nm as determined

by measurement of 100 particles from such images. PEA-UCNPs were then covalently immobilized on

aldehyde modified paper via reduction amination between amine groups available on PEA-UCNPs and

aldehyde moieties on the surface. Immobilization was confirmed by comparing the luminescence

intensity obtained from 1.5 mg mL-1 PEA-UCNPs spotted on aldehyde modified paper with that from

paper that was treated with sodium cyanoborohydride reagent in the absence of aldehyde groups. The

latter showed minimal retention of PEA-UCNPs with a luminescence intensity that was only 4.9 ± 0.3 %

9

compared to the aldehyde functionalized paper. A similar immobilization strategy for UCNPs has been

previously shown to demonstrate excellent stability on aldehyde functionalized glass surfaces [5]. In

addition to providing a larger surface area compared to glass surfaces, the use of paper provides a

greener chemistry platform. Glass surface derivitization depends on the use of hundreds of milliliters of

organic solvent such as toluene, methanol, dichloromethane, and ethyl ether[5]. However, only

microliter volumes of an aqueous lithium chloride and sodium (meta)periodate solution were used to

obtain aldehyde functionalized paper.

While similar LRET assays are based on the adsorption of UCNPs on paper[17], we have observed that

covalent conjugation is more reproducible and robust. Covalently immobilized UCNP systems provided

emission intensity with a standard deviation of about 8% for 5 spots treated with PEA-UCNPs, compared

to a standard deviation close to 18% for UCNPs adsorbed on paper. Adsorption procedures involve

drying the paper, which can cause the UCNPs to aggregate. UCNPs in the centre of an aggregated group

of particles cannot participate in LRET and this results in lower LRET ratios[17]. It was thus important to

ensure that the spots on the paper remained hydrated during covalent UCNP immobilization and

reaction time was limited to 10 minutes to prevent drying. Hydrated spots on aldehyde functionalized

glass slides showed a uniform monolayer of UCNPs with no aggregation[5].

3.2 Characterization of QD-DNA conjugates:

A well established method was used to quantify the amount of hexahistidine modified DNA on the

QDs[20]. An extinction coefficient of 210,000 L mol−1 cm−1 was used for the QDs[27]. It was determined

that incubating the QDs with approximately 6, 12, 24 and 36x DNA yielded an average of 6.0 ± 0.2, 12.0

± 0.3, 19.3 ± 0.3 and 19.8 ± 0.3 DNA strands per QD (Figure S2). Hexahistidine has been shown to have a

very high affinity for ZnS overcoated QDs with excellent control over QD: DNA ratio[24]. While a

10

maximum of approximately 20 DNA strands were immobilized on the surface of the QDs, the use of salt

aging can allow the immobilization of up to 40 DNA strands[16] . Thus there remains surface area

available for all of the QD-DNA conjugate ratios reported herein.

The DNA:QD conjugates were then incubated with 15x excess hexahistidine modified PEG 6000. This

material is able to coordinate to the surfaces of QDs to create a PEG coating around the QDs.

Confirmation was required that DNA molecules would not desorb or be displaced by the hexahistidine.

The gel image in Figure 2 confirms that an increasing DNA:QD ratio is maintained after the addition of

the PEG and the absorbance spectra of the filtrates from 100 KDa centrifugal filter reduced to 40µL

showed no desorption of DNA (Figure S3). While PEG only decorated QDs showed minimal

electrophoretic mobility in the gel image in Figure 2, wells (iii) to (vi) show an increasing electrophoretic

mobility, which is proportional to the number of nucleic acids on the surface of the QD. The increase in

mobility is a result of the negative charges associated with the phosphate backbone of the nucleic acids

on the surface of the QDs[27]. The similar electrophoretic mobilities of the QDs in wells (v) and (vi)

indicate that both samples have similar DNA:QD ratios and confirm the similar ratios of 19.3:1 and

19.8:1 obtained in the absorbance experiments. Thus, QDs were incubated with up to 24x excess DNA

only in subsequent experiments.

11

(a) (b)

Figure 2: (a) Image of 2 % agarose gel developed in 0.5xTBE at 5.7 V cm -1 for (i) GSH-QDs, (ii) PEG-GSH-QDs, PEG-GSH-QDs incubated with (iii) 6x, (iv) 12x, (v) 24x and (vi) 36x excess hexahistidine modified DNA. (b) line profiles of lanes (ii) to (vi).

3.3 Optimization of DNA:QD ratio for hybridization assay

The immobilized UCNPs were used as energy donors in an LRET-based sandwich assay for detection of

unlabeled target oligonucleotide. A gene fragment from HPRT1, a housekeeping gene found in

mammalian cells, was selected as a generic target representative of a typical segment of DNA.

Housekeeping genes are important in maintaining basic cell function and are expected to maintain a

constant expression level in healthy individuals. The detection of housekeeping genes is routinely used

for control and calibration in many biotechnological applications and genomic studies[28]. The

immobilized DNA probes on the UCNPs served to capture a portion of the HPRT1 gene fragments.

Unhybridized portions of the gene fragments extending from the surface then captured reporter

oligonucleotides that were immobilized on QDs. Theoretical calculation indicates that the QDs will be

within approximately 15 nm of the UCNPs if the hybrids are perpendicular to the surface and are fully

12

extended. The 3-dimensional nature of paper allows neighboring UCNPs to be closer to the captured

QD, especially upon contraction of the paper matrix once it is dry[16].

Bednarkiewicz et al. reported a decrease in the lifetime of UCNPs in the presence of QDs indicating that

QD emission can be sensitized by LRET[29]. In another study, a UCNP and QD LRET pair was used for the

detection of thrombin on glass substrates, however it showed modest performance due to low numbers

of donors associated with a monolayer of UCNPs on the surface[5]. Herein, this LRET pair is used in the

3-dimensional matrix of paper to demonstrate use for DNA detection that can offer up to a 10-fold

enhancement in sensitivity[16].

The spectral overlap between NaYF4: 0.5% Tm3+, 30% Yb3+/NaYF4 core/shell UCNPs and 575 QDs is shown

in Figure 3. The QD emission was collected in the absence of any UCNP luminescence using a band pass

filter that transmitted wavelengths between 565 and 585 nm, as seen in Figure 3. This effectively

eliminated background signal from the luminescence of the donor in the acceptor channel. On the

contrary, donor photoluminescence leakage into the acceptor optical channel is typical in UCNP-dye[5]

and QD-dye[15] donor-acceptor systems and can lead to low signal-to-background ratios and poor

sensitivity.

13

Figure 3: An overlay of UCNP emission (black), 575 QD absorbance (green) and emission (red) and 570/20 nm filter transmittance (blue) spectra demonstrating the passage of QD fluorescence and the blocking of UCNP luminescence. The data for transmittance and the optical density of the filter can be found in Figure S4.

14

It is important to ensure minimal non-specific adsorption of the QD reporters on the paper to obtain

reproducible and reliable results. The immobilized PEA-UCNPs were decorated with NHS-PEG4-biotin

because PEG is well known to minimize non-specific adsorption [30]. The NHS-PEG4-biotin solution was

dissolved in a low volume to minimize NHS hydrolysis in water and was pipetted immediately to ensure

the presence of unhydrolized NHS groups that were available to react with the amine groups on the

UCNPs. The biotin on the PEG served to capture avidin, which was then used for immobilization of

biotinylated probe oligonucleotide. The paper was treated with a 20 µM solution of a non-

complementary unlabelled oligonucleotide to ameliorate non-specific binding of the surface with the

nucleic acids conjugated to the QD reporters. Even with such blocking of the surface, it was noticed that

the use of GSH-QDs decorated with 6.0 ± 0.2 hexahistidine modified nucleic acids resulted in non-

specific adsorption that was proportional to the QD concentration (Figure 4). The interaction between

the QDs and the surface in the absence of HPRT1 target was hypothesized to be driven by the GSH

molecules that coated the QD. The GSH-QDs were then conjugated with PEG to prevent the GSH

molecules from interacting with the surface of the paper. Figure 4 shows very minimal adsorption of the

PEG stabilized GSH-QDs that were decorated with oligonucleotides compared to the DNA decorated

GSH-QDs with no PEG in the absence of HPRT1 target. Background signals in the QD optical channel

were subtracted before computing the LRET ratio.

15

Figure 4: Minimal retention of PEG stabilized oligonucleotide decorated GSH-QDs (second bar of each pair) compared to oligonucleotide decorated GSH-QDs (first bar of each pair) is observed even with increasing concentration of conjugated QDs in the absence of HPRT1 target.

Figure S5 demonstrates that non-specific adsorption of the PEG stabilized DNA decorated GSH-QDs on

paper is independent of the number of DNA strands on the surface of the QD. Figure S5 also

demonstrates that the same LRET ratio is obtained for the detection of 0.5 µM HPRT1 target regardless

of the amount of HPRT1 reporters on the QDs when 0.5 µM QDs were used. This suggests that the

amount of DNA around a QD does not influence the LRET ratio obtained for the same HPRT1 target

concentration. We explored this possibility further and varied the target concentrations to ensure that

the observation was not due to the relative concentration of HPRT1 targets to QD reporters. The target

concentrations were chosen to be 0.01, 0.1. 0.25 and 0.5 µM, while the QD concentration was chosen to

be 0.25 µM. Figure 5 demonstrates that the LRET ratio obtained for HPRT1 target concentrations less

than, equal to and greater than the QD concentration was independent of the amount of HPRT1

reporter on the QD. This suggests that the extent of capture of a reporter on the surface is not primarily

dependent of the number of reporters on each QD.

To explore whether the rate of hybridization was limited by the diffusion of the QDs, kinetic experiments

for the detection of 0.5 µM of HPRT1 target using different reporter DNA:QD ratios were performed.

Additionally, experiments were completed using Cy5 labeled reporter oligonucleotide in place of QDs

that were coated with reporter. The results shown in Figure 6 suggest similar kinetics for all of the

reporter DNA:QD ratios. Hybridization with Cy5 labeled reporter DNA was noted to be substantially

faster than hybridization with QDs that were coated with reporter. Experiments that used the Cy5

labeled oligonucleotides achieved 95% of the full signal within the first minute of reaction, compared to

three minutes for the QDs that were decorated with reporter. Because all of the various ratios of

16

DNA:QD conjugates showed similar kinetics, the QDs were incubated with 6x excess reporter

oligonucleotide in subsequent experiments, and signal was measured after 5 minutes of reaction time.

Figure 5: Normalized LRET ratio obtained for increasing concentrations of the HPRT1 target for QD: HPRT1 reporter 1:6.0 (grey), 1:12.0 (blue) and 1:19.3 (orange).

Figure 6: Kinetic curves fit to exponential functions for the hybridization of reporter DNA:QD 6.0:1 (green), 12.0:1 (grey and 19.3:1 (blue)and Cy5 labeled reporter DNA (black). The overall trends show that conjugating the reporter on the QD slows signal generation, and that the reaction time is insensitive to the DNA:QD ratio for the conditions used in these experiments.

3.4 Hybridization assays using clean samples

17

Figure 7 shows a calibration curve for the detection of HPRT1 gene fragment with concentrations

ranging from 5 nM to 2 µM using a DNA-QD concentration of 2 µM. The response curve demonstrated a

limit of detection of 13.31 ± 0.08 fmol and dynamic range spanning nearly 3 orders of magnitude. The

limit of detection was calculated as three standard deviations above the LRET ratio in the presence of a

non-complementary target and QD reporters. The limit of linearity of 6 pmol was constrained by the

amount of QDs to HPRT1 target.

Figure 7 Response curve for the detection of HPRT1 target.

The calibration curve in Figure 7 shows improved analytical performance compared to similar paper-

based assays that used UCNPs as donors and molecular dyes as acceptors[17]. This is primarily

attributed to QDs being superior acceptors since they exhibit better extinction coefficients and broader

absorption spectra compared to molecular fluorophores. While the reported assay in the literature

made use of DNA targets that were tagged with fluorescent dyes, our assay for unlabeled target exhibits

an order of magnitude better sensitivity and an order of magnitude wider dynamic range, in addition to

a 2.5 fold improvement in LOD.

18

Figure 8: Assay response for 2 µM HPRT1 target (grey) and 2 µM HPRT1 1BPM (purple) in the presence of 0, 5 and 10% formamide demonstrating the selectivity of the assay.

The selectivity of the assay was evaluated by comparing the assay response from a fully complementary

target to one base pair mismatched target, and the resulting data is presented in Figure 8. Contrast

ratios of fully complementary target to one base pair mismatched target were 2.9, 9.8 and 60.5 in the

presence of 0%, 5% and 10% formamide in the wash buffer, respectively. These values compare

favorably to the values of 1.2, 1.9 and 3.9 reported for an UCNP-molecular dye system [17]. The latter

ratios represent relatively low contrast and it is likely that fouling of the paper by the dye-labelled

oligonucleotide sequences was only partially addressed by the introduction of formamide. The ability to

decorate QDs with PEG to minimize non-specific adsorption demonstrates another advantage of use of

QDs as LRET acceptors.

3.5 Hybridization assay in goat serum

The analytical performance of the assay in 90% serum was negatively impacted in comparison to that in

buffer, but still achieved a limit of detection of 24 fmol and a dynamic range spanning over 2 orders of

19

magnitude (Figure 9). The LRET system in serum still showed figures of merit that were superior in

comparison to UCNP-molecular dye systems in buffer, where the limit of detection for the latter was 34

fmol and the dynamic range was less than 2 orders of magnitude[17]. The selectivity of the assay was

also evaluated in 90% serum (Figure S6) and contrast ratios of fully complementary target to one base

pair mismatched target were 3.3, 4.6 and 30.6 in the presence of 0%, 5% and 10% formamide in the

wash buffer, respectively.

Figure 9: Response curve for the detection of HPRT1 target in serum using a concentration of 2 µM of QD reporters.

4. Conclusions

We have demonstrated the first use of covalent immobilization of UCNPs on paper for development of

an LRET-based assay for unlabeled oligonucleotide target. The UCNP-QD LRET pair allowed for the

collection of QD photoluminescence in the absence of donor background owing to the narrow and well

defined emission bands of both nanoparticles. The use of PEG decorated QDs as LRET acceptors

demonstrated improved sensitivity, limit of detection, dynamic range and selectivity compared to

20

analogous assays that have used molecular fluorophores as acceptors. For samples in buffer, the assay

was optimized to achieve a limit of detection of 13 fmol, a dynamic range spanning nearly 3 orders of

magnitude and a contrast ratio of fully complementary target to one base pair mismatched target of

60.5 : 1. The assay showed good performance for determination of target oligonucleotide in 90% serum

samples, with a limit of detection of 24 fmol and a dynamic range spanning over 2 orders of magnitude.

The kinetics of hybridization were found to be diffusion limited and independent of the amount of

reporter DNA on the QDs, and signal could be collected within 3 minutes.

Acknowledgments:

We thank Dr. Sreekumari Nair for obtaining TEM images and Omair M. Noor for useful discussion. We

gratefully acknowledge the Natural Sciences and Engineering Research Council of Canada (NSERC) for

financial support of this research. S.D. is thankful to the Ontario Ministry of Training, College and

Universities (MTCU) for provision of an Ontario Graduate Scholarship (OGS), and U.U. is grateful to

NSERC for a graduate fellowship.

21

References

[1] M.V. DaCosta, S. Doughan, Y. Han, U.J. Krull, Lanthanide upconversion nanoparticles and applications in bioassays and bioimaging: A review, Anal. Chim. Acta 832 (2014) 1-33.[2] T. Rantanen, M.L. Järvenpää, J. Vuojola, R. Arppe, K. Kuningas, T. Soukka, Upconverting phosphors in a dual-parameter LRET-based hybridization assay, Analyst 134 (2009) 1713-1716.[3] M. Kumar, P. Zhang, Highly sensitive and selective label-free optical detection of DNA hybridization based on photon upconverting nanoparticles, Langmuir 25 (2009) 6024-6027.[4] J. Liu, J. Cheng, Y. Zhang, Upconversion nanoparticle based LRET system for sensitive detection of MRSA DNA sequence, Biosens. and Bioelectron. 43 (2013) 252-256.[5] S. Doughan, Y. Han, U. Uddayasankar, U.J. Krull, Solid-phase covalent immobilization of upconverting nanoparticles for biosensing by luminescence resonance energy transfer, ACS Appl. Mater. Interfaces 6 (2014) 14061-14068.[6] Y. Wang, L. Bao, Z. Liu, D.W. Pang, Aptamer biosensor based on fluorescence resonance energy transfer from upconverting phosphors to carbon nanoparticles for thrombin detection in human plasma, Anal. Chem. 83 (2011) 8130-8137.[7] H. Chen, F. Yuan, S. Wang, J. Xu, Y. Zhang, L. Wang, Aptamer-based sensing for thrombin in red region via fluorescence resonant energy transfer between NaYF4: Yb,Er upconversion nanoparticles and gold nanorods, Biosens. Bioelectron. 48 (2013) 19-25.[8] H. Li, L. Wang, NaYF4:Yb3+/Er3+ nanoparticle-based upconversion luminescence resonance energy transfer sensor for mercury(ii) quantification, Analyst 138 (2013) 1589-1595.[9] C. Li, J. Liu, S. Alonso, F. Li, Y. Zhang, Upconversion nanoparticles for sensitive and in-depth detection of Cu 2+ ions, Nanoscale 4 (2012) 6065-6071.[10] B. Liu, H. Tan, Y. Chen, Upconversion nanoparticle-based fluorescence resonance energy transfer assay for Cr(III) ions in urine, Anal. Chim. Acta 761 (2013) 178-185.[11] P. Kannan, F. Abdul Rahim, R. Chen, X. Teng, L. Huang, H. Sun, D.H. Kim, Au nanorod decoration on NaYF4:Yb/Tm nanoparticles for enhanced emission and wavelength-dependent biomolecular sensing, ACS Appl. Mater. Interfaces 5 (2013) 3508-3513.[12] Y. Yuan, Z. Liu, An effective approach to enhanced energy-transfer efficiency from up-converting phosphors and increased assay sensitivity, Chem. Comm. 48 (2012) 7510-7512.[13] T. Rantanen, H. Päkkilä, L. Jämsen, K. Kuningas, T. Ukonaho, T. Lövgren, T. Soukka, Tandem dye acceptor used to enhance upconversion fluorescence resonance energy transfer in homogeneous assays, Anal. Chem. 79 (2007) 6312-6318.[14] A.W. Martinez, S.T. Phillips, G.M. Whitesides, E. Carrilho, Diagnostics for the developing world: Microfluidic paper-based analytical devices, Anal. Chem. 82 (2010) 3-10.[15] M.O. Noor, U.J. Krull, Paper-based solid-phase multiplexed nucleic acid hybridization assay with tunable dynamic range using immobilized quantum dots as donors in fluorescence resonance energy transfer, Anal. Chem. 85 (2013) 7502-7511.[16] M.O. Noor, U.J. Krull, Camera-based ratiometric fluorescence transduction of nucleic acid hybridization with reagentless signal amplification on a paper-based platform using immobilized quantum dots as donors, Anal. Chem. 86 (2014) 10331-10339.

22

[17] F. Zhou, M.O. Noor, U.J. Krull, Luminescence resonance energy transfer-based nucleic acid hybridization assay on cellulose paper with upconverting phosphor as donors, Anal. Chem. 86 (2014) 2719-2726.[18] F. Zhou, U.J. Krull, Spectrally Matched Duplexed Nucleic Acid Bioassay Using Two-Colors from a Single Form of Upconversion Nanoparticle, Anal. Chem. 86 (2014) 10932-10939.[19] M.O. Noor, E. Petryayeva, A.J. Tavares, U. Uddayasankar, W.R. Algar, U.J. Krull, Building from the "Ground" Up: Developing interfacial chemistry for solid-phase nucleic acid hybridization assays based on quantum dots and fluorescence resonance energy transfer, Coord. Chem. Rev. 263-264 (2014) 25-52.[20] R. Freeman, X. Liu, I. Willner, Chemiluminescent and chemiluminescence resonance energy transfer (CRET) detection of DNA, metal ions, and aptamer-substrate complexes using hemin/G-quadruplexes and CdSe/ZnS quantum dots, J. Am. Chem. Soc. 133 (2011) 11597-11604.[21] M. Kumar, D. Zhang, D. Broyles, S.K. Deo, A rapid, sensitive, and selective bioluminescence resonance energy transfer (BRET)-based nucleic acid sensing system, Biosens. Bioelectron. 30 (2011) 133-139.[22] W.R. Algar, D. Wegner, A.L. Huston, J.B. Blanco-Canosa, M.H. Stewart, A. Armstrong, P.E. Dawson, N. Hildebrandt, I.L. Medintz, Quantum dots as simultaneous acceptors and donors in time-gated Förster resonance energy transfer relays: Characterization and biosensing, J. Am. Chem. Soc. 134 (2012) 1876-1891.[23] H.S. Qian, Y. Zhang, Synthesis of hexagonal-phase core-shell NaYF4 nanocrystals with tunable upconversion fluorescence, Langmuir 24 (2008) 12123-12125.[24] I.L. Medintz, L. Berti, T. Pons, A.F. Grimes, D.S. English, A. Alessandrini, P. Facci, H. Mattoussi, A Reactive Peptidic Linker for Self-Assembling Hybrid Quantum Dot−DNA Bioconjugates, Nano Lett. 7 (2007) 1741-1748.[25] U. Uddayasankar, Z. Zhang, R.T. Shergill, C.C. Gradinaru, U.J. Krull, Isolation of monovalent quantum dot-nucleic acid conjugates using magnetic beads, Bioconj. Chem. 25 (2014) 1342-1350.[26] J.-C. Boyer, M.-P. Manseau, J.I. Murray, F.C.J.M. van Veggel, Surface Modification of Upconverting NaYF4 Nanoparticles with PEG−Phosphate Ligands for NIR (800 nm) Biolabeling within the Biological Window, Langmuir 26 (2009) 1157-1164.[27] U. Uddayasankar, R.T. Shergill, U.J. Krull, Evaluation of Nanoparticle–Ligand Distributions To Determine Nanoparticle Concentration, Anal. Chem. 87 (2014) 1297-1305.[28] E. Eisenberg, E.Y. Levanon, Human housekeeping genes, revisited, Trends in Genetics 29 (2014) 569-574.[29] A. Bednarkiewicz, M. Nyk, M. Samoc, W. Strek, Up-conversion FRET from Er3+/Yb3+:NaYF4 nanophosphor to CdSe quantum dots, J. Phys. Chem. C 114 (2010) 17535-17541.[30] C. Blaszykowski, S. Sheikh, M. Thompson, Surface chemistry to minimize fouling from blood-based fluids, Chem. Soc. Rev. 41 (2012) 5599-5612.

23

Figure Captions:

Figure 2: The strategy for the detection of target DNA. Probe oligonucleotide is conjugated to an UCNP,

and the UCNP is immobilized on a paper substrate. Target oligonucleotide serves to bridge probe

oligonucleotide on the UCNP and reporter oligonucleotide on the QD. Excitation of the UCNP at 980 nm

provides for luminescence that excites the QD. The paper substrate is prepared to localize 3 reaction

spots that are defined by wax printing. The orange, black and green nucleic acid strands represent the

probe, target and reporter strands, respectively.

Figure 2: (a) Image of 2 % agarose gel developed in 0.5xTBE at 5.7 V cm -1 for (i) GSH-QDs, (ii) PEG-GSH-

QDs, PEG-GSH-QDs incubated with (iii) 6x, (iv) 12x, (v) 24x and (vi) 36x excess hexahistidine modified

DNA. (b) line profiles of lanes (ii) to (vi).

Figure 3: An overlay of UCNP emission (black), 575 QD absorbance (green) and emission (red) and

570/20 nm filter transmittance (blue) spectra demonstrating the passage of QD fluorescence and the

blocking of UCNP luminescence. The data for transmittance and the optical density of the filter can be

found in Figure S4.

Figure 4: Minimal retention of PEG stabilized oligonucleotide decorated GSH-QDs (second bar of each

pair) compared to oligonucleotide decorated GSH-QDs (first bar of each pair) is observed even with

increasing concentration of conjugated QDs in the absence of HPRT1 target.

Figure 5: Normalized LRET ratio obtained for increasing concentrations of the HPRT1 target for QD:

HPRT1 reporter 1:6.0 (grey), 1:12.0 (blue) and 1:19.3 (orange).

Figure 6: Kinetic curves fit to exponential functions for the hybridization of reporter DNA:QD 6.0:1

(green), 12.0:1 (grey and 19.3:1 (blue)and Cy5 labeled reporter DNA (black). The overall trends show

that conjugating the reporter on the QD slows signal generation, and that the reaction time is insensitive

to the DNA:QD ratio for the conditions used in these experiments.

Figure 7: Response curve for the detection of HPRT1 target.

Figure 8: Assay response for 2 µM HPRT1 target (grey) and 2 µM HPRT1 1BPM (purple) in the presence

of 0, 5 and 10% formamide demonstrating the selectivity of the assay.

Figure 9: Response curve for the detection of HPRT1 target in serum using a concentration of 2 µM of

QD reporters.

24