Development of an Oligonucleotide Functionalized Hydrogel

May 8, 2009 - Integrated on a High Resolution Interferometric Readout. Platform as a ... gated due to the wide application of these sensors. The major...
0 downloads 0 Views 790KB Size
Download PDF
Biomacromolecules 2009, 10, 1619–1626

1619

Development of an Oligonucleotide Functionalized Hydrogel Integrated on a High Resolution Interferometric Readout Platform as a Label-Free Macromolecule Sensing Device Sven Tierney and Bjørn Torger Stokke* Biophysics and Medical Technology, Department of Physics, The Norwegian University of Science and Technology, NTNU, NO-7491 Trondheim, Norway Received February 19, 2009; Revised Manuscript Received April 13, 2009

Development of an oligonucleotide functionalized hydrogel integrated on a high resolution interferometric readout platform capable of determining changes in optical length of the hydrogel with 2 nm resolution is described. The hydrogels were designed with hybridized dioligonucleotides grafted to the polymer network making up a network junction point in addition to the covalent cross-links. The hybridized dioligonucleotide network junctions were made with a 10 basepair complementary region flanked by additional basepairs that could aid in destabilizing the junction points in competitive displacement hybridization by the added probe oligonucleotides. The probe oligonucleotide destabilizing the junction point thus induces swelling of the functionalized hydrogel that is sensitive to the concentration of the probe, the sequence, and matching length between the probe and sensing oligonucleotide. This design yields a molecular amplification of the change in the optical length of the gel at least 5-fold compared to a hydrogel where sensing functionality is based on hybridization with a grafted oligonucleotide that is not a part of a network junction. Concentration sensitivity applied for specific label-free detection of oligonucleotide is estimated to be in the nanomolar region. Applications of the resulting oligonucleotide imprinted hydrogel for label-free sensing of probe oligonucleotide sequences or taking advantage of the oligonucleotide sequences designed with aptamer functionalities for determination of other types of molecules are discussed.

1. Introduction Development of DNA sensors has been intensively investigated due to the wide application of these sensors. The majority of DNA-sensor systems rely on labeling of DNA with, for example, fluorescent1-4 or enzymatic5,6 markers. In spite of the high sensitivity of these techniques, there has been an increasing interest in label-free detection of specific nucleic acids sequences. Such migration toward label-free detection is motivated by circumventing the labeling procedure and thereby also reducing the cost of the sensors, and a possibly more rapid detection method. Several read-out techniques such as surface plasmon resonance,7-9 voltametric,10 and QCM9,11 have been utilized. So far there has been limited investigation into the use of functionalized hydrogels for DNA sensing. In their early work, Nagahara and Matsuda12 reported on the synthesis of oligonucleotides grafted on succinimido-copolymers yielding an overall comblike structure, and the absorbance properties of the oligonucleotide branches occurring on hybridization. The grafted oligonucleotides were mechanistic in supporting gel formation by hybridized sequences making up junction zones and thereby interconnecting the polymer chains. Further investigations on oligonucleotide functionalized hydrogels have been carried out by Yurke et al.13,14 and Maeda et al.15 In all cases it was shown that the properties of the samples were affected when exposed to an aqueous solution containing oligonucleotide with a complementary sequence. In this paper we have employed a scheme where an acrylamide-based hydrogel modified with acrylamide-DNA monomers (acrydite16) incorporated. The overall aim is to prepare a hydrogel with equilibrium swelling volume that * To whom correspondence should be addressed: E-mail: bjorn.stokke@ ntnu.no.

depends on the concentration and sequence of an oligonucleotide added to the solution. The main strategy consisted of incorporating two complementary strands in the gel as a physical crosslink in addition to the covalent cross-links. When a free oligonucleotide with a complementary sequence to that grafted to the network is added to solution, competitive displacement17,18 takes place within the hybridized dsDNA oligomer constituting the physical cross-link and thereby disconnects the polymer backbones at this point (Figure 1). In such a way, dissociation of the connectivity mediated by the imprinted dioligonucleotide allows the hydrogel to adjust to a new equilibrium swelling volume. The swelling degree of the hydrogels was measured by an optical interferometric technique (Figure 1a). The high resolution of this readout platform being approximately 2 nm for changes in optical length, ∆lopt, of the hydrogel and detection rate of 1 Hz provides a highly sensitive determination of the swelling of the functionalized hydrogels and thereby its response to oligonucleotides sequences and concentration. In addition to being sensitive to complementary sequences for the displacement hybridization, selection of the grafted oligonucleotide sequence with aptamer functionality enhances the ability to apply this type of hydrogel material for specific label free detection of a broad range of biological relevant molecules.19,20 Such a strategy was reported to induce a sol-gel transition in oligonucleotide grafted on a polymer backbone,21 or employing surface plasmon resonance imaging,22 and other readout platforms.23,24

2. Materials and Methods Oligonucleotide grafted acrylamide-bis-acrylamide hydrogels were prepared at the end of an optical fiber as previously described.25 In brief, the surface functionalization with an outermost layer of meth-

10.1021/bm900218c CCC: $40.75  2009 American Chemical Society Published on Web 05/08/2009

1620

Biomacromolecules, Vol. 10, No. 6, 2009

Tierney and Stokke Table 1. Sequence of Oligonucleotides Integrated as Sensing (S) and Blocking (B) Units in the Acrydite Functionalized Polyacrylamide Hydrogels and Probe (P) Oligonucleotides Used To Test the Hydrogel Swelling Responsea S1 S2 B1 P1 P2 P3 P4 P5 P6 R

Figure 1. Schematic illustration of the principle for determination of changes in the optical length of a hemispherical biosensisitive hydrogel attached to the end of an optical fiber (a) and oligonucleotide cross-linking sensitive hydrogel (b-f). Incident light (wavelength range λa-λb from 1530 to 1560 nm) directed through an optical fiber reflected at the fiber-gel and gel-solution interfaces, R1 and R2, yields a reflected interference wave with phase depending on the optical path length depending on the hydrogel swelling. The hemispheral hydrogel at the end of the optical fiber has a length of the order 45-55 µm (optical length of 60-75 µm) in this case (a). Sensing (blue) and blocking (red) oligonucleotides are grafted in their hybridized state and by their 5′-acrydite functionalized ends to the polymer backbone and, thereby, making up additional cross-links to the covalent ones (black dots; b,d). Probe nucleotides (green; e) added to the solution acts by competetive displacement hybridization with the sensing oligonucleotide, expelling the blocking nucleotide, and thereby destabilize the junction functionality of the hybridized dioligonucleotide (c,f). Disruption of dioligonucleotide junction functionality leads to an altered length of the gel depending on the properties of the probe oligonucleotide such as concentration, basepair match for blocking length (10) and between the probe and sensing oligonucleotide (10-20; b,c).

acrylate groups to covalently link the gel to the fiber consisted of cleaning the fiber in ethanol followed by NaOH, rinsing, and activation by 0.01 M HCl before soaking in a solution of of 3-(trimethoxysilyl) propyl methacrylate (0.02 M, nitrogen-purged MQ water, pH ) 3.5) for 1 h chemically binding the methacrylate groups to the fiber tip. The resulting surface-bound methacrylate groups were used to covalently link the hydrogel to the surface. The hydrogels were synthesized using acrylamide (acrylamide, 99%, Sigma), bis(N,Nmethylene-bisacrylamide, 99%+, Acros organics) as the covalent crosslinker, and hybridized oligonucleotides (5′-acrydite, Integrated DNA Technologies) with each oligonucleotide strand connected through their 5′-end to topologically separated network chains. The oligonucleotide sequences were designed by a script written in Matlab. The sequence of the sensing oligonucleotide S1 was created by choosing bases at random. The blocking oligonucleotide B1 was designed as a totally 20 bp oligomer with a 10 bp sequence complementary to S1 thus supporting a 10 bp dsDNA hybridized structure within S1-B1. Bases 11-20 of B1 were therefore predetermined to be complementary to that of S1, whereas base 1-10 were then created at random. The same strategy was employed for the probeDNAs (designated P1-P6, 20-10 bp match, 35 basepairs total). The random sequence oligonucleotide was created by the same strategy as S1. The script also checked the match length in case of randomly generated extra bp matches. The various employed oligonucleotides are listed in Table 1. The melting temperatures of the hybridized dsDNA were estimated at the used solution conditions by applying the procedure as described by Owczarzy et al.26

5′-TGC ACC CAC AAT GGA GCC AG-3′ 5′-ACC CGA TAG TAG GTT GAG CG-3′ 5′-CCC CGA TGT GCT GGC TCC AT-3′ 5′-CAA AAC TCC CAA TTG CTG GCT CCA TTG TGG GTG CA-3′ 5′-TCG CGG GAG GAA GCC GCG TTC TGG CTC CAT TGT GG-3′ 5′-CTT AAC TCC TGG CTC CAT TGA CTG GCT GGC GCT TT-3′ 5′-ATC CCT TAC TGG CTC CTT TGG GCA CGG ACC GCA TA-3′ (1 mismatch) 5′-GTA GGT TGC TGG CTC CAT AGC GAT AGC TTG TAA AC-3′ 5′-ATA GCC GCT CAA CCT ACT ATT TGT AAA CAC AGT GAC TGG CTC CAT TGT GG-3′ 5′-AAA GTT GAG CGT GTC ACT CCG AAC GAG ACA CTG AC-3′

a S1 and B1 are 10 base pair complementary. Oligonucleotides P1, P2, P3, P4 and P5 are complementary to S1 with bp match of 20, 15, 12, 11 with 1 mismatch and 10. P6 is complementary with S1 and S2 (15 match with both). R is a random sequence (4 match). In the fluorescence experiments S1 was tagged with 5′-Cy3 whereas P1 was tagged with 3′Iowa-black.

Appropriate amounts of AAM (12 wt % stock solution, 0.3 mol % bis), bis (0.46 wt %), acrydite, and free oligonucleotides (if used) were mixed together with the buffer (45 mM NaCl, 10 mM Tris, 1 mM EDTA, yielding an ionic strength I ∼ 56 mM) and yielding 10 wt % AMM. Gels with varying amounts of the S1 and B1 oligonuleotides were synthesized, as were gels with the S1 and S2. In the latter case, this was conducted both with and without oligonucleotide P6. The pregel solution was left for at least 3 h prior to inducing the crosslinking-copolymerization thus allowing the hybridization of the S1 and B1 oligonucloetides. Subsequently, the pregel solution was deposited at the end of methacrylate functionalized optical fiber via a pipet in a squalane (99%, Aldrich) solution saturated with a photoinitiator (hydroxycyclohexylphenylketone,99%, Aldrich). This solution was cured by UV light (Dymax Bluewave 50). The process is explained in more detail elsewhere.25 When measuring, the fibers were placed in glass tubes with an inner diameter of 1 mm. The gel was situated at one end of the tube there was a cavity of approximately 100 µL. A total of 30 µL of buffer solution (pH 7.4 10 mM Tris (Sigma Aldrich, 99.8%+), 1 mM EDTA (Sigma Aldrich, 99%), 150 mM NaCl (Sds, 99%)) was injected in the cavity and the tube was sealed with Teflon tape. The tube was located in a thermostatted water bath ((0.1 °C) until thermal equilibrium was reached. The fiber/tube was taken out of the water bath and preheated stock solutions of the oligonucleotide probes (20 µL) were added before resealing and re-immersing in the water bath. It should be stressed here that, when adding the probe solution, the buffer/probe solution was mixed by suctioning the liquid in the pipet several times. The complete procedure took approximately 60s. Changes in optical length, ∆lopt, were determined for at least 1 h following exposure to the solution with the probes. The data was logged 20 s after reimmersion into the water bath due to an initial temperature effect. After an experiment, the gel was placed in a DMSO solution for 15 minutes to denature the probe-S1, that is, to wash out the probes. Subsequently, the gel was immersed in the buffer solution until constant phase. It was assumed that the hybridized pair of the S1 and B1 oligonucleotides were regenerated thus regenerating the cross-linking functionality when the constant phase was attained. This assumption is based on regeneration of a nearly identical absolute optical length and reproducibility of the kinetics of swelling following this washing procedure (see below). Parameter ∆lopt of the gel was determined from the interference between the waves reflected at the fiber-gel and gel-solution interfaces, yielding a change in the phase and using a light source

Label-Free Macromolecule Sensing Device (1530-1560 nm) as described.25,27 The observed ∆lopt may originate both from changes of the physical length (l) and the refractive index (ngel) of the gel.27 The 2 nm detection limit ∆lopt for a 50 µm hydrogel correspond to a detection limit of changes in the refractive index of 5.3 × 10-5, assuming changes only in n. Eventual uneven changes of the refractive index along the lightpath trough the material will give rise to detectable changes based on differences of the integral of ngel along the lightpath, that is, the technique provides the net change, and does not provide information about eventual inhomgeneous changes of ngel. The kinetics of the displacement hybridization of the free S1-B1 pair by oligonucleotide P1, was carried out in the same buffer solution employing Cy3-5′-oligonucleotide (Integrated DNA Technologies) and Iowa Black-3′-oligonucleotide (Integrated DNA Technologies). The dye (Cy3) was conjugated to the sensing strand, S1fluo, whereas the quencher was conjugated to one of the probes (20 bp match), P1quench. Fluorescence spectroscopy was carried out, employing an excitation wavelength λ ) 550 nm, and detecting at the emission wavelength 564 nm (Spex Spectramate Fluorolog, Spex industries, Inc., NJ). Appropriate amounts of the oligonucleotides were dissolved to the desired concentration in the Tris-HCl buffer (pH ) 7.4, I ) 160 mM), and the S1fluo and B1 were mixed. The S1fluo - B1 solution was inserted into the spectrometer and equilibrated for 2 min followed by addition of P1quench and the fluorescence intensity was monitored. The concentration of S1fluo and B1 were constant for every measurement, whereas the displacement of B1 on S1fluo by P1quench was monitored at various P1quench concentrations.

3. Results and Discussion 3.1. Strategy of Preparing Oligonucleotide Sensitive Polymer Gel. An overall strategy where dsDNA segments connected with either ssDNA to topologically separated parts of the polymer network can be implemented using two principally different strategies. The function of the incorporated ssDNA is to engage in hybridization supporting a network junction sensitive to ssDNA in solution. In the first of these strategies (Figure 1), each of two, at least partial, complementary ssDNA were allowed to hybridize and incorporated into the gel during the cross-linking-polymerization reaction. The sensitivity toward ssDNA in solution is mediated by the competetive displacement of the network cross-linking hybridized pair, resulting in a hybridized pair and a ssDNA oligomer connected to separate network chains. The competitive displacement induced by the oligonucleotide added to the solution, with the network junction dioligonucleotide depends on the concentration and bp matches (Figure 1). This process reduces the crosslinking density depending on the bp match conditions and concentration of the probe oligonucleotides added to the solution. There is a 2-fold rationale underlying this strategy: First the DNA monomers in the pregel solution hybridize prior to synthesis, thus molecular imprinting similar to reported for other biomolecular pairs28,29 is attained. Second the basepair match requirements serves to screen unwanted reactions with free DNA, the free DNA has to competitively displace the bp match within the gel. The latter process requires a larger basepair match between the probe oligonucleotide and the oligonucleotide grafted to the network, than the base-pair match between the two complementary strands within the gel. The grafted oligonucleotide, referred to as the sensing strand (S), forms a 10 bp hybridized duplex with the other grafted nucleotide, the blocking strand (B). The functionalized hydrogels were exposed to several probe oligonucleotides (Table 1) and at varying concentrations. The probe oligonucleotides were designed with a sequence with 20 to 10 bp complementarity to the sensing strand of the gel. All the probe oligonucleotides

Biomacromolecules, Vol. 10, No. 6, 2009

1621

were 35 bp and the oligonucleotides grafted on the polymer network chain (acrydite monomers) were 20 bps. Furthermore, various molar concentrations of the acrydite monomers incorporated in the hydrogel were used to explore the relative swelling response of the hydrogel matrix when exposed to various concentrations of ssDNA. An alternative approach utilizes two noncomplementary oligonucleotides incorporated into the polymer chain of the hydrogel where cross-links can be formed due to hybridization of both oligonucleotides with one probe.30 The expected hybridization between the probe oligonucleotide and the two complementary network grafted oligonucleotides is foreseen to increase the cross-linking density thus affecting the hydrogel equilibrium swelling. However, at the rather low oligonucleotide grafting density used here, this strategy did not appear to support such an effect, at variance with other reports.13,15,30 The present apparent lack of success to induce such oligonucleotide crosslinking may arise from a large separation between the grafted oligonucleotides at the used low grafting density. The basic design employing the hybrized oligonucleotide pair grafted to the network (Figure 1), making up probe ssDNA sensitive crosslinks is therefore the only design pursued in the following. The bp match within the hybridized segments constituting the DNA sensitive junction zones should support a stable network junction in absence of probe oligonucleotide while at the same time be susceptible to competitive hybridization toward the probe oligonucleotide. The gel with 0.4 mol % oligomer junction zones corresponds to approximately 4 mM in the gel, provided 100% conversion. In an aqueous buffer solution with ionic strength of 160 mM, the 10 bp dsDNA used as junction zones is estimated to have a melting temperature Tm ) 68.9 ( 2 °C.26 A Tm ) 54.53 ( 2 °C is estimated for a conversion of 1%. For the experiments performed here, the maximum temperature was 37 °C, which is well below these predicted Tm. The selection of a 10 bp match region for the hybridized pair incorporated as a cross-link appear, from the point of view of required stability, to have a sufficient overlap region to avoid spontaneous dissociation at physiological temperature. An additional design parameter of the gel is the relative amount of acrydite within the gel. Various molar fractions of acrydite were tested and there appeared to be an optimal range. As an example, Figure 2 shows the response of two gels, incorporating 0.4 and 0.7 mol % acrydite, respectively, to 20 µM P1 (Table 1). The P1 oligonucleotide has a 20 bp match with the S1 incorporated in the gel. The data show that the gel with 0.4 mol % hybridized oligonucleotide network cross-links reacts much more rapidly with 20 µM P1 and is exhausted after 2.0 × 103 s, whereas the 0.7 mol % acrydite hydrogel was not exhausted within the time limit of the experiment. Additional combinations of the ratio between the covalent and dsDNA supported cross-linkers were investigated by varying the molar concentrations of Bis and S1-B1 (results not shown). The results based on evaluation of the initial swelling kinetics of these biofunctionalized gels when exposed to P1 at various concentrations, showed that the gels with 0.4 mol % acrydite apparently yielded the best sensitivity at 0.6 mol % Bis. The sizes of these gels are larger than the previous gels we have investigated using this fiber-optic readout-platform.25,27 This is due to the higher degree of ionization and lower degree of crosslinker in the present gels. The acrydite groups have 20 basepairs implying 20 charges per grafted molecule. A gel with 0.4 mol % dioligonucleotide supported cross-links has the same charge density as a 16 mol % acrylic acid gel. The reason for the higher sensitivity for the 0.4 mol % S1-B1 hydrogels, may arise from

1622

Biomacromolecules, Vol. 10, No. 6, 2009

Tierney and Stokke

associated with adoption of a new ∆lopt that principally may occur following 2.1 × 103 s of 20 µM P1 exposure provided a cross-link-insensitive state. The increase of ∆lopt versus time of the 0.4 mol % S1-B1 gel at various step changes of the concentration of P1 shows a kinetic of swelling depending monotonically on the P1 concentration. Exposing these oligonucleotide functionalized hydrogels to P1 appear to yield an initial ∆lopt that increases approximately linearly with time. The observed dependence of the initial slope, ∆lopt/∆t, on the concentration of the probe DNA (Figure 3b) suggests that the technique can be used as a basis for quantitatively determining a probe DNA in solution. Note, however, that this requires that the bp match of the probe and network immobilized sensing oligonucleotide is known.

Figure 2. Change in the optical length of 0.4 and 0.7 mol % S1-B1 dioligonucleotide functionalized, 0.6 mol % bis, 10 wt % acrylamide hydrogels to 20 µM oligonucleotide P1 added in solution. Measurements were carried out at pH 7.4, I ) 160 mM, and temperature 37 °C. Large symbols are shown for every 200 data point and the expriments were carried out two times for both concentrations of S1-B1 incorporated in the gels.

the large charge density of the network; with a highly charged gel the swelling degree increase might be quite small for the highly swollen gels. Differences in steric constraints (see below) between the 0.7 and 0.4 mol % acrydite gel may also be a factor affecting the relative swelling response of these hydrogels. The data (Figure 2) also show that the second exposure to the same concentration of the probe oligonucleotides following the washing procedure involving DMSO yields almost identical development of ∆lopt versus time. In view of the sensitivity of development of ∆lopt to the concentration of the acrydite (Figure 2), the reproducibility following the washing procedure for a given P1-B1 concentration suggests regeneration of the hybridized S1-B1 pair following the washing procedure. Similar reproducibilities following the washing, including DMSO observed here, also in the data presented in the following, indicate that regeneration of hybridized S1-B1 pair cross-link density are general for the developed hydrogels. 3.2. Sensitivity to Concentration of Probe Oligonucleotides. A change of 18 µm of the optical length was determined when exposing the 0.4 mol % acrydite gels to aqueous solutions with 20 µM P1 that is designed with a 20 bp match with the S1 of the junction (Figure 3a). The new equilibrium swelling state at 20 µM P1was reached after about 2.1 × 103 s (Figure 3a). As a first approximation, the new equilibrium ∆lopt is suggested to arise from completion of the displacement hybridization reaction between the 20 µM P1and the S1-B1 pair. A P1 concentration lowered to 2 µM yielded a slower change in the swelling of the gel, and the data show an almost linear increase of ∆lopt during the 4.8 × 103 s data were collected, with no indication of a new equilibrium swelling state in this period. Similarly, ∆lopt determined for exposure to a lower concentrations of P1 does not indicate a new equilibrium swelling states within the first 4.8 × 103 s. It therefore appears that the initial kinetics of ∆lopt is more relevant as the readout parameter than the new equilibrium swelling state. Such an analysis is not influenced by an eventual incompleted displacement hybridization reaction

This experimental series was performed down to a P1 concentration of 500 nM. The total change in ∆lopt of about 3000 nm within 4.8 × 103 s (Figure 3) for the particular DNA functionalized gel exposed to 500 nM P1 is clearly discernible, and it appears that the resolution in concentration for this gel is approaching 1/5 of 500 nM. However, it should be noted that assuming that the experimentally determined uncertainty of 2 nm, as induced by P1 within 120 s, would correspond to a concentration of 6-7 nM. For the swelling experiments induced by a 500 nM P1 concentration, an overlaid apparent oscillation of ∆lopt with time was observed. Such oscillations for changes in swelling in equilibrium gel volume from one equilibrium state to the other were not observed where changes in ionic strength or pH were used to change the ∆lopt of cationic gels27 or various carbohydrates using boronic acid as the sensing moiety in the gel.25 In addition to the various concentrations of P1, an oligonucleotide with a random basepair sequence, R, was also tested. This strand has a 4 match with S1, and as expected due to the shielding by the 10 bp S1-B1 hybridized sequence, no reaction is seen. If anything, there is a slight decrease in the gel length similar to that reported by others.31 Hydrogels with only the sensing oligonucleotide, that is, without the blocking oligonucleotide were additionally prepared to explore eventual amplification of the signal associated with the design using displacement hybridization destabilizing a junction point occurring on molecular recognition. The results show that probe oligonucleuotides hybridizing with the grafted sensing oligonucleotides induces a change in the optical length increasing with time (Figure 3c) similar to the functional hydrogels including the blocking strands, but to a significant lesser degree (compare ∆lopt in Figure 3a,c at equivalent P1 concentration). The amplification of the molecular recognition signal at 0.5 and 2 µM P1 are about 5.5 times for the hydrogel design utilizing destabilization of a junction compared to the hydrogel response based only on recognition with a grafted S1. A change in the charge density associated with the hybridization is a possible molecular mechanism leading to the observed change in ∆lopt of the gels (Figure 3c). Such a hypothesis relies on a change in charge density for the grafted appendices occurring on hybridization altering the parameters of the Donnan equilibrium, but not the cross-linking density, in the standard theoretical approach describing swelling of hydrogels.32,33 For the hydrogels designed with a destabilizing junction occurring concomitant with the recognition of the probe oligonucleotide, the cross-linking density will also be a primary affected parameter in the theory for hydrogel swelling coupling the molecular interaction to the swelling behavior.

Label-Free Macromolecule Sensing Device

Biomacromolecules, Vol. 10, No. 6, 2009

1623

Figure 3. Optical length of the 0.4 mol % S1-B1 dioligonucleotide, 0.6 mol % bis, 10 wt % acrylamide hydrogel versus time following addition of P1 olionucleotide (Table 1) to the aqueous solution at concentrations 20, 2, and 1 µM and 500 nM. The measurements were carried out twice at pH 7.4, I ) 160 mM, and temperature 37 °C except for the measurement at 500 nM, which was carried out in triplicate (a). Initial rate of change of optical length with time ∆lopt/∆t vs concentration of P1 oligonucleotide in solution used to induce swelling of 0.4 mol % S1-B1 dioligonucleotide, 0.6 mol % bis, 10 wt % acrylamide hydrogel gel (b). The data of ∆lopt/∆t is shown as the mean ( SD of the slopes estimated from the data in (a) using linear regression and ∆t up to 500 and 1000 s and also forcing the fit through origo. Optical length of the 0.4 mol % S1, 0.6 mol % bis, 10 wt % acrylamide hydrogel gel vs time following addition of P1 olionucleotide to the aqueous solution at concentrations 2 µM and 500 nM (c). The measurements were carried out twice at pH 7.4, I )160 mM, and 37 °C. The large symbols in Figures 3a and c are shown for every 200 data point collected, and the intermediate apparent lines depict individual data points.

Figure 4. Optical length of the 0.4 mol % S1-B1 dioligonucleotide, 0.6 mol % bis, 10 wt % acrylamide hydrogel vs time following addition of probe oligonucleotides with 20 (P1), 15 (P2), 12 (P3), and 10 (P5) bp match (Table 1) toward the S1 oligonucleotide of the gel (a). Data obtained for probe oligonucleotide P4 with one bp mismatch relative to P3 is also shown. The measurements were carried out twice at 2 µM probe oligonucleotides at pH 7.4, I ) 160 mM, and 37 °C. Initial rate of change of optical length with time ∆lopt/∆t vs bp match length between the probe oligonucleotides and S1 of the gels characterized in (a) (b). The data of ∆lopt/∆t is estimated from the data in (a) according to Figure 3b. The large symbols are shown for every 200 data points collected, and the intermediate apparent lines depict individual data points.

3.3. Sensitivity to bp Match Length and Sequence of Probe Oligonucleotides. The number of bps in the complementary region between the probe and sensing oligonucleotides within the displacement reaction was investigated using probe oligonucleotides with sequence match of 10, 12, 15, and 20 bp within oligomers of total 35 bp (Figure 1, Table 1). Additionally, a probe with 12 bp sequence match, but 1 bp mismatch, was also explored (P4, Table 1). The effect of bp match determined at a concentration of 2 µM of the various probe oligomers (Figure 4) shows that the largest bp match toward the grafted S1 induces the most rapid increase in gel swelling. The P2 probe with a 15 bp match induces a fairly rapid change in the gel swelling volume, but not as fast as the P1. Both of these probes support a much longer bp match with S1 than the 10 bp match

of the S1-B1 pair grafted in the network. The observed difference in the kinetics is thus expected to arise from a larger equilibrium constant of the P1-S1 pair (20 bp match) than the P2-S1 pair (15 bp match). The initial differences in ∆lopt/∆t induced by the oligonucleotide probes with various bp complementarities to S1 were estimated by linear regression for the data obtained at 2 µM probe concentration. The data (Figure 4b) show that ∆lopt/∆t using the P2 (15 bp match with S1) is about 70% of the P1 (20 bp match with S1). The 12 bp match probe displays an initial rate of change of about 20% of the 20 bp match probe. The trend clearly shows that the ∆lopt/∆t approaches zero when the bp match length with S1 is reduced toward the bp match of the S1-B1 pair. There is, however, a small change of ∆lopt observed also for a bp match length of 10

1624

Biomacromolecules, Vol. 10, No. 6, 2009

Figure 5. Optical length of the 0.4 mol % S1-B1 dioligonucleotide, 0.6 mol % bis, 10 wt % acrylamide hydrogel vs time following addition of 2 µM probe oligonucleotides with 20 (P1) bp match (Table 1) toward the S1 oligonucleotide of the gel at pH 7.4, I ) 160 mM and temperatures 25, 30, and 37 °C. The large symbols depicts every 200 data points collected (two series), and the intermediate apparent lines depict individual data points.

bp with S1 suggesting that there is a displacement equilibration based on a dynamic exchange process taking place. The stability of the S1-B1 pair relative to the probe for the various number of bp matches is thus a key factor contributing to reaction with the probe strand and, thereby, a central design parameter to control the performance of these materials. Both the oligonucleotides P3 and P4 also react with the gel (Figure 4). The 1-mismatch, located at basepair nr 9 on P4, does however show a slower reaction than P3, indicating that it is possible to discriminate a single bp mismatch, provided the total concentration of the probe DNA and overall length of the complementary strand is known. Increasing the temperature when determine the probe nucleotide induced change in ∆lopt yields a more rapid change the larger the temperature (Figure 5). The trends for the initial change in ∆lopt shows a tendency toward an equilibrium plateau for the experiments at 25 °C, whereas there is a more rapid change of ∆lopt at 30 and 37 °C, and no indication of a longterm plateau within the first 4.8 × 103 s for the experiments at 37 °C. The more rapid displacement reaction in the strand exchange process probably underlying the increase of ∆lopt/ ∆t arises both from the general increased thermal activition the higher the temperature and the reduced temperature span between the actual temperature and the Tm of the S1-B1dioligonucleotide. 3.4. Swelling Kinetics. The above data indicate that a new equilibrium of ∆lopt of the hydrogels are obtained only for the 20 µM P1 after about 2.1 × 103s (Figure 3a). Other conditions did not reveal a clear sign of a new equilibrium of ∆lopt within 1 h. Due to this rather slow response, further investigations were carried out to explore possible rate determining steps. The following three mechanisms can be identified to possibly affect the overall sensor response: (1) diffusion of the probe oligonucleotides in the gel network to reach the sensing-blocking dioligonucleotide, (2) the displacement of the blocking oligonucleotide by the probe oliginucleotide, and (3) swelling of the gel network. The kinetics of the swelling of the gel network was characterized by exposing a 0.40 mol % S1-B1 gel to stepwise increases of ionic strength (∼155 mM). The data (not shown) revealed that the new plateau of ∆lopt was reached within seconds. This is in line with our previous report on ionic strength induced swelling of an cationic gel of the actual size.27 Note that time

Tierney and Stokke

Figure 6. Normalized fluorescence intensity vs time following addition oligonucleotide P1 labeled with Iowa black (P1quench) to a solution of 0.5 µM of the hybrized pair of fluorescence labeled (Cy3) oligonucleotide S1 (S1fluo) and oligonucleotide B1. The experiments were carried out in aqueous buffer (I ) 160 mM) and using concentrations of P1quench as indicated. The large symbols depicts every 100 data point collected (two series), and the intermediate apparent lines depict individual data points.

constant for adoption of the new plateau of ∆lopt for ionic strength induced re-equilibration in the order of a few seconds is much smaller the time constants for the probe oligonucleotide induced changes. This indicates that the kinetic of gel network swelling is not the rate determining step of the oligonucleotide sensing hydrogel material studied here. The kinetics of displacement hybridization of the S1-B1 pair with the P1 oligonucleotide was determined for the free oligonucleotides using fluorescence labeled S1 (S1fluo), and quencher labeled P1 (P1quench). The kinetics of fluorescence quenching of solutions of the hydridized S1fluo-B1 oligonucleotide pair by the P1quench oligonucleotide at various concentrations are shown in Figure 6. The data show that the normalized fluorescence intensity, If, is quenched within ∼9.0 × 102 s for a stoichiometric ratio between S1fluo and P1quench, at 0.5 µM S1fluo. Decreasing the concentration of the P1quench below stoichiometric relative to S1fluo reduces the initial rate of the decrease in If, but the new level of If appears to be reached within a similar period as for the 0.5 µM P1quench. Compared to the time constant of seconds for the kinetics of swelling of the hydrogels induced by changes in ionic strength, the time constant for quencing of hydridized S1fluo-B1 oligonucleotide pair by the P1quench is much larger. This set of data thus confirms that the strand exchange process is substantially slower than the swelling kinetics of the gel network of the actual size. The diffusion of the probe within the gel network is another potential limiting factor for the kinetics of the response of the hydrogel material. A characteristic limiting diffusion constant, Dlim for the oligonucleotide probe diffusion to be the rate determining factor for the hydrogel response is estimated based on the physical dimensions of the hemispheral gel of radius 50 µm and applying the relation 〈r2〉 ) 6Dt, where 〈r2〉 is the mean square diffusion distance of a molecule with diffusion constant D within the time t. For a typical time constant of 5.0 × 102 s for near completion of the displacement hybridization in solution (Figure 6), Dlim is estimated to 0.83 × 10-8 cm2 s-1. The present estimate is in line with the estimated charactestic time for diffusion of 10 s for short DNA fragments in the gels of similar size of a cell 20-30 µm.34 The unconstrained diffusion coefficient of ssDNA in aqueous solution for bp > 10 has been estimated to35

Label-Free Macromolecule Sensing Device

D ) 7.38 × 10-6(bp)-0.539

Biomacromolecules, Vol. 10, No. 6, 2009

(1)

where D is the normalized diffusion coefficient to standard solvent conditions at 20 °C and given in cm2 s-1. For the probes used here (bp ) 35), the value of D(bp ) 35) ) 110 × 10-8 cm2 s-1 and the estimated Dlim above indicates that oligonucloetide diffusion is not the rate limiting factor for the displacement hybridization reaction determined for the free oligonucleotides. The ratio between the free solution diffusion constant and that in gels have been reported to about 0.1 for globular proteins and PEG in acrylamide gels with volume fraction about 0.1.36 Note that the scaling coefficient -0.539 in eq 1 indicates that the ssDNA adopts a random coil type conformation in solution and, therefore, that the effect of the volume fraction of the polymer component of the hydrogel can be expected to follow the same scaling behavior for the oligonucleotides probes employed here as that reported for PEG. Thus, the mobility of the probe oligonucleotides in the hydrogel sensor does not appear to be a limiting factor for the kinetics of the response. Steric constraints imposed by the proximity of the polymer backbone and the grafted oligonucleotide may be a candidate obstacle contributing to the difference in the kinetics observed for the displacement hybridization in solution and the kinetics of oligonucleotide induced swelling of the hydrogels. Similarly, hybridization of DNA at a solid/liquid interface is often slower than in solution.37 Nevertheless, the sensor material does display characteristic response depending on the probe concentration and sequence that reflect these parameters that can be tuned by the selection of various design parameters.

4. Conclusions In this work we have developed a hydrogel material consisting of a synthetic polymer network with hybridized olgionucleotides grafted to the network chain, thereby yielding dioligonucleotide supported junction points contributing to the elastic response of the material. The junction functionality of the network dioligonucleotides was destabilized by competing oligonucleotides added to the aqueous solution. The associated decrease of total number of junctions of the network resulted in an increase in the equilibrium swelling of the hydrogel material. The changes in the hydrogel swelling were determined by the preparation of the gel material as a hemisphere grafted to the surface of an optical fiber. Changes in the optical length of the hydrogel with a resolution of 2 nm support the application of the developed material in a label free sensing scheme of oligonucleotide with high resolution. The dioligonucleotide supported junctions are designed with a sensing nucleotide sequence that is completely or partially blocked by a complementary nucleotide sequence. The probe oligonucleotide hybridizes with the network grafted sensing nucleotide, displaces the blocking nucleotide and, thereby, suspends the junction functionality. The material is designed with large flexibility of the sequence length to be probed. When this oligonucleotide functionalized hydrogel material is used and when a user selected oligonucleotide sequence is employed, label-free oligonucleotide sensing is demonstrated with concentration-sensitive response for a 20 bp hybridization sequence. A probe oligonucleotide concentration detection of 500 nM is demonstrated, and the technique appears to support concentration detection in the range of a few nM for the employed network composition. For the 10 bp blocking sequence, the rate of change of the optical length of the hybrid material was shown to be dependent on the number of basepair match between the probe and sensing nucloetide. This supports label-free detection

1625

of single-nucleotide replacements in the case that the sequence and concentration are known. Although supporting label-free detection of oligonucleotides, decoupling of the concentration and bp match length awaits further developments. Selection of the grafted dioligonucleotides with one of the oligonucleotide possessing aptamer functionality extends the applicability of the designed hydrogels to sense numerous other molecules. Examples of this include aptamer functionality supporting the specific sensing of, for example, human thrombin,21 histone H4,24 cellobiose,38 or other molecules.39 In designing biosensors utilizing the aptamer functionality of the grafted sequences, the present strategy offers versatility in being able to tune sensitivity and kinetics by selection of network parameters and bp match parameters. Acknowledgment. This work was supported by the strategic university program in medical technology at NTNU, Project Number 154080, supported by the Norwegian Research Council. The authors gratefully acknowledge the experimental assistance and discussions with Thor Bernt Melø (Department of Physics, The Norwegian University of Science and Technology) for the fluorescence measurements.

References and Notes (1) Piunno, P. A. E.; Krull, U. J.; Hudson, R. H. E.; Damha, M. J.; Cohen, H. Anal. Chem. 1995, 67, 2635–2643. (2) Duveneck, G. L.; Pawlak, M.; Neuschafer, D.; Bar, E.; Budach, W.; Pieles, U. Sens. Actuators, B 1997, 38, 88–95. (3) Ferguson, J. A.; Boles, T. C.; Adams, C. P.; Walt, D. R. Nat. Biotechnol. 1996, 14, 1681–1684. (4) Walt, D. R. Science 2000, 287, 451–452. (5) Hillier, S. C.; Flower, S. E.; Frost, C. G.; Jenkins, A. T. A.; Keay, R.; Braven, H.; Clarkson, J. Electrochem. Commun. 2004, 6, 1227–1232. (6) Pividori, M. I.; Merkoci, A.; Alegret, S. Biosens. Bioelectron. 2001, 16, 1133–1142. (7) Jordan, C. E.; Frutos, A. G.; Thiel, A. J.; Corn, R. M. Anal. Chem. 1997, 69, 4939–4947. (8) Hu, W. P.; Chen, S. J.; Huang, K. T.; Hsu, J. H.; Chen, W. Y.; Chang, G. L.; Lai, K. A. Biosens. Bioelectron. 2004, 19, 1465–1471. (9) Su, X. D.; Wu, Y. J.; Knoll, W. Biosens. Bioelectron. 2005, 21, 719– 726. (10) Panke, O.; Kirbs, A.; Lisdat, F. Biosens. Bioelectron. 2007, 22, 2656– 2662. (11) Caruso, F.; Rodda, E.; Furlong, D. F.; Niikura, K.; Okahata, Y. Anal. Chem. 1997, 69, 2043–2049. (12) Nagahara, S.; Matsuda, T. Polym. Gels Network 1996, 4, 111–127. (13) Yurke, B.; Lin, D. C.; Langrana, N. A. Use of DNA nanodevices in modulating the mechanical properties of polyacrylamide gels. DNA Computing; Springer-Verlag: Berlin, 2006; pp 417-426. (14) Liedl, T.; Dietz, H.; Yurke, B.; Simmel, F. Small 2007, 3, 1688–1693. (15) Murakami, Y.; Maeda, M. Biomacromolecules 2005, 6, 2927–2929. (16) Rehman, F. N.; Audeh, M.; Abrams, E. S.; Hammond, P. W.; Kenney, M.; Boles, T. C. Nucleic Acids Res. 1999, 27, 649–655. (17) Li, Q. Q.; Luan, G. Y.; Guo, Q. P.; Liang, J. X. Nucleic Acids Res. 2002, 30 (e5), 1–9. (18) Reynaldo, L. P.; Vologodskii, A. V.; Neri, B. P.; Lyamichev, V. I. J. Mol. Biol. 2000, 297, 511–520. (19) Bunka, D. H. J.; Stockley, P. G. Nature ReV. Microbiol. 2006, 4, 588– 596. (20) Carothers, J. M.; Oestreich, S. C.; Szostak, J. W. J. Am. Chem. Soc. 2006, 128, 7929–7937. (21) Yang, H. H.; Liu, H. P.; Kang, H. Z.; Tan, W. H. J. Am. Chem. Soc. 2008, 130, 6320–6321. (22) Li, Y.; Lee, H. J.; Corn, R. M. Anal. Chem. 2007, 79, 1082–1088. (23) Cruz-Aguado, J. A.; Penner, G. Anal. Chem. 2008, 80, 8853–8855. (24) Lin, L.; Hom, D.; Lindsay, S. M.; Chaput, J. C. J. Am. Chem. Soc. 2007, 129, 14568–14569. (25) Tierney, S.; Volden, S.; Stokke, B. T. Biosens. Bioelectron. 2009, 24, 2034–2039. (26) Owczarzy, R.; You, Y.; Moreira, B. G.; Manthey, J. A.; Huang, L. Y.; Behlke, M. A.; Walder, J. A. Biochemistry 2004, 43, 3537–3554. (27) Tierney, S.; Hjelme, D. R.; Stokke, B. T. Anal. Chem. 2008, 80, 5086– 5093.

1626

Biomacromolecules, Vol. 10, No. 6, 2009

(28) Miyata, T.; Asami, N.; Uragami, T. Nature 1999, 399, 766–769. (29) Miyata, T.; Jige, M.; Nakaminami, T.; Uragami, T. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 1190–1193. (30) Jiang, F. X.; Yurke, B.; Firestein, B. L.; Langrana, N. A. Ann. Biomed. Eng. 2008, 36, 1565–1579. (31) Murakami, Y.; Maeda, M. Macromolecules 2005, 38, 1535–1537. (32) Treloar, L. R. G. The Physics of Rubber Elasticity; Clarendon Press: Oxford, 1975. (33) Shibayama, M.; Tanaka, T. AdV. Polym. Sci. 1993, 109, 1–62. (34) Livshits, M. A.; Mirzabekov, A. D. Biophys. J. 1996, 71, 2795–2801.

Tierney and Stokke (35) Stellwagen, E.; Lu, Y. J.; Stellwagen, N. C. Biochemistry 2003, 42, 11745–11750. (36) Tong, J.; Anderson, J. L. Biophys. J. 1996, 70, 1505–1513. (37) Teles, F. R. R.; Fonseca, L. R. Talanta 2008, 77, 606–623. (38) Yang, Q.; Goldstein, I. J.; Mei, H. Y.; Engelke, D. R. Proc. Natl. Acad. Sci. U.S.A. 1998, 95, 5462–5467. (39) Torres-Chavolla, E.; Alocilja, E. C. Biosens. Bioelectron. 2009, doi: 10.1016/j.bios.2008.11.010.

BM900218C