Effects of Oligonucleotide Immobilization Density on Selectivity of

4984

Langmuir 2000, 16, 4984-4992

Effects of Oligonucleotide Immobilization Density on Selectivity of Quantitative Transduction of Hybridization of Immobilized DNA James H. Watterson,†,‡ Paul A. E. Piunno,†,‡ Christopher C. Wust,§ and Ulrich J. Krull*,‡ Chemical Sensors Group, Department of Chemistry, University of Toronto at Mississauga, 3359 Mississauga Road North, Mississauga, Ontario, Canada, L5L 1C6, and FONA Technologies Inc., 855 Matheson Boulevard East, Unit #14, Mississauga, Ontario, Canada, L4W 4L6 Received November 17, 1999 Immobilized single-stranded DNA (ssDNA) can be used as a selective “reagent” to bind complementary nucleic acids for applications including detection of pathogenic organisms and genetic mutations. The density of ssDNA on a surface will determine nearest neighbor interactions, surface interactions, and charge density due to ionizable phosphate groups. This may result in a local ionic strength, pH, and dielectric constant at the surface that is substantially different from that in bulk electrolyte solution. It is the local conditions that influence the thermodynamics of hybridization, and this can be studied by the melt temperature (Tm) of double-stranded DNA (dsDNA). Organosilane chemistry has been used to covalently immobilize hexaethylene glycol linkers and to control the subsequent density of dT20 that was prepared by automated synthesis. Fiber-optic biosensors based on fused silica optical fibers that were coated with DNA were used in a total internal reflection fluorescence instrument to determine Tm from the dissociation of duplexes of mixtures of fluorescein-labeled and unlabeled dA20 and d(A9GA10). Each thermal denaturation of dsDNA at the surface of the optical fibers was accompanied by a 2-3-fold reduction in standard enthalpy change, relative to values determined for denaturation in bulk solution. The experimental results suggest that the thermodynamic stability of duplexes that are immobilized on a surface is dependent on the density of immobilized DNA. Additionally, the deviation in Tm arising as a result of the presence of a centrally located single base-pair mismatch was significantly larger for thermal denaturation occurring at the surface of the optical fibers (∆Tm ) 6-10 °C) relative to that observed in bulk solution (∆Tm ) 3.8-6.1 °C). These results suggest that hybridization at an interface occurs in a significantly different physical environment in comparison to hybridization in bulk solution, and that surface density can be tuned to design analytical figures of merit.

Introduction Newer techniques that are used in the detection of microorganisms include immunoassay 1,2 and immunosensor3-5 techniques, which tend to rely on protein binding as the means of molecular “recognition”, as well as those which make use of nucleic acid hybridization,6-11 as the basis for selective recognition. Biochips and biosensors can make use of immobilized nucleic acids to provide for † The authors of this manuscript wish it to be known that P.A.E.P and J.H.W. have participated equally in the experimentation and preparation of this manuscript and should, in their opinion, be considered joint-first-authors. * Author to whom correspondence should be addressed. ‡ University of Toronto at Mississauga. § FONA Technologies Inc.

(1) Hage, D. S. Anal. Chem. 1993, 65, 420R. (2) Seare, N. J. Immunoassay Techniques. In Chemical Sensors Edmonds, T. E., Ed.; Chapman and Hall: New York, 1988; 155. (3) Blonder, R.; Katz, E.; Cohen, Y.; Itzhak, N.; Riklin, A.; Willner, I. Anal. Chem. 1996, 68, 3151. (4) Granzow, R.; Reed, R. Biotechnology 1992, 10, 390. (5) Ko¨nig, B.; Gra¨tzel, M. Anal. Chim. Acta 1995, 309, 19. (6) Millan, K. M.; Saraullo, A.; Mikkelsen, S. M. Anal. Chem. 1994, 66, 2943. (7) Wang, J.; Bollo, S.; Lopez Paz, J. L.; Sahlin, E.; Mukherjee, B. Anal. Chem. 1999, 71, 1910. (8) Su, H.; Kallury, K. M. R.; Thompson, M.; Roach, A. Anal. Chem. 1994, 66, 769. (9) Caruso, F.; Rodda, E.; Furlong, D. N.; Niikura, K.; Okahata, Y. Anal. Chem. 1997, 69, 2043. (10) Abel, A. P.; Weller, M. G.; Duveneck, G. L.; Ehrat, M.; Widmer, H. M. Anal. Chem. 1996, 68, 2905. (11) Piunno, P. A. E.; Krull, U. J.; Hudson, R. H. E.; Damha, M. J.; Cohen, H. Anal. Chem. 1995, 67, 2635.

selective binding interactions. These systems are attractive since the selectivity of nucleic acid binding interactions can be quite high and the advent of solid phase nucleic acid synthesis has allowed for relatively simple nucleic acid immobilization to combine the selective sensing chemistry with the transduction element. The thermal stability of the double-helix state is often measured by means of the thermal denaturation temperature, Tm, which is defined as the temperature at which half of all duplexes originally formed are denatured into the single-stranded state.12 The thermal stability of doublestranded DNA (dsDNA) is dependent on base composition, as illustrated by differences of Tm for oligomers of the same size with different G-C base pair content. Basepair mismatches can reduce the observed Tm. The magnitude of this reduction depends on the number of mismatches and length of the duplex in question. In duplexes that are 150 or more nucleotides in length, each 1% increase in base-pair mismatch content has been reported to reduce the observed Tm by 0.5-1.4 °C.13 Each base-pair mismatch has been reported to reduce the observed Tm by up to 5 °C for smaller oligomers that are 14-20 nucleotides in length.14 The observed Tm is also a function of total strand concentration and ionic strength of the surrounding solution. Some stabilization imparted (12) Nelson, J. W.; Martin, F. H.; Tinoco, I., Jr. Biopolymers 1981, 20, 2509. (13) Bonner, T. I.; Brenner, D. T.; Neufeld, B. R.; Britten, R. J. J. Mol. Biol. 1973, 81, 123. (14) Wallace, R. B.; Shaffer, J.; Murphy, R. F.; Bonner, J.; Hirose, T.; Itakura, K. Nucl. Acids Res. 1979, 6, 3543.

10.1021/la991508m CCC: $19.00 © 2000 American Chemical Society Published on Web 04/21/2000

Oligonucleotide Immobilization Density Effects

by higher salt concentrations may also be attributed to increased screening of repulsive effects of the polyanionic backbone of the nucleic acid oligomers.15 Plots of Tm versus log [Na+] generally show a slope of 18-22 °C, for oligonucleotides that are 16 or more nucleotides in length.16 The value of Tm is also weakly dependent upon strand concentration, whereas the values of the enthalpy and entropy changes accompanying denaturation are not. A dependence of the thermodynamic parameters and the observed Tm at higher strand concentrations (above 75 µM) has been attributed to strand aggregation, which is thought to preferentially stabilize dsDNA relative to ssDNA which did not display the tendency to aggregate in the ultracentrifugation study carried out by Tinoco et al.12 The use of nucleic acid oligomers as selective binding elements often involves immobilization of the oligomers at solid interfaces. Immobilized oligomers at a solid surface can experience significantly different environments from those experienced in bulk solution, which can add further complexity to the control of selectivity and linearity of response with the concentration of complementary oligonucleotide sequences. The immobilization of singlestranded DNA (ssDNA) onto the surface of a solid substrate affects the charge density at the surface. The density of immobilized ssDNA may therefore affect the electrostatic properties of the surface in such a way as to affect local ionic strength and the extent to which the immobilized oligomers interact with the surface of the solid substrate, which in turn may determine the extent to which the immobilized ssDNA is available for hybridization. These factors may influence the Tm behavior of the hybrids formed at the surface. This has consequences to issues of selectivity and the extent of hybridization, and it is likely that the orientation of the immobilized ssDNA may affect the kinetics of hybridization.17 A more rigorous characterization of the surface environment of biochips and biosensors that utilize immobilized nucleic acids is required to elucidate the effects of interfacial immobilization on selective nucleic acid binding and thermodynamic stability. The present work has investigated the impact of the density of immobilized ssDNA on the observed Tm, and has investigated the effects of salt concentration and the presence of single base-pair mismatches on the thermodynamics of hybridization. Experimental Section Chemicals. Solvents were obtained from BDH (Toronto, ON) as reagent grade and were further purified or dried, when necessary, by standard distillation methods. Reagent grade salts were purchased from BDH (Toronto, ON). DNA synthesis reagents were from Dalton Chemical Laboratories Inc. (Toronto, ON). Anhydrous acetonitrile (Dalton) was dried by distillation from P2O5 prior to receipt and was further distilled from calcium hydride under a dry argon atmosphere prior to use. Tetrahydrofuran (BDH) was first dried over CaH2, filtered, and finally distilled immediately prior to use from sodium metal (Aldrich)/ benzophenone (Aldrich). Sterile water for use on its own and with hybridization buffer was produced with the water first double-distilled in glass and then subsequently treated with diethyl pyrocarbonate (Aldrich) and sterilized by autoclave. Molecular biology grade polyacrylamide gel electrophoresis reagents and apparatus were obtained from Bio-Rad (Hercules, CA). Silica gel (Toronto Research Chemicals, Toronto, ON) that was used for purification had a particle size of 30-70 µm. (15) Puglisi, J. D.; Tinoco, I., Jr. Methods Enzymol. 1989, 180, 304. (16) Breslauer, K. J. In Methods in Molecular Biology, Vol. 26: Protocols for Oligonucleotide Conjugates; Agrawal, S., Ed.; Humana Press: Totowa, NJ, 1994; p 347. (17) Su, H.; Williams, P.; Thompson, M. Anal. Chem. 1995, 67, 1010.

Langmuir, Vol. 16, No. 11, 2000 4985 Preparation of Optical Fiber Segments. Fused silica optical fibers of 400 µm core diameter (3M Powercore Series Optical Fiber, FT-400-URT or FP-400-UHT) were acquired from Thor Labs Inc., Newton, NJ. The polymeric outer cladding was removed mechanically by means of a fiber-stripping tool also obtained through Thor Labs, Inc. Removal of the outer cladding exposed the inner cladding layer that coated the core of fused silica. Individual sensor elements were then made by cutting optical fiber pieces 48 mm in length with a fiber-scoring device. The fiber was cleanly scored by rotating a diamond pencil about the optical fiber. Removal of the top portion of the scored fiber from the remainder of the optical fiber secured in the pin-chuck then yielded cylindrical optical fiber segments with clean, flat termini as evidenced by visual inspection of the termini at 40× magnification. The fused silica fiber segments and controlled pore glass (CPG) (CPG Inc., Lincoln Park, NJ) that were used as solid substrates for automated DNA synthesis were cleaned prior to modification of the surface according to the two-stage method of Kern and Puotinen.18 The first stage consisted of immersing the solid substrates in a 1:1:5 (v/v) solution of 30% ammonium hydroxide/ 30% hydrogen peroxide/water and gently agitating at 80 °C for 5 min. In the second stage, the substrates were then recovered, thoroughly washed with sterile water, and then gently agitated in a solution of 1:1:5 (v/v) concentrated HCl/30% hydrogen peroxide/water for 5 min at 80 °C. The substrates were then recovered and washed with successive 100 mL portions of water, methanol, dichloromethane, and diethyl ether. The substrates were then dried under vacuum and stored in vacuo and over P2O5 until required. Surface Modification of Solid Substrates: Functionalization of Substrates with 3-Glycidoxypropyltrimethoxysilane (GOPS). The cleaned solid substrates were suspended in an anhydrous solution of xylene/3-glycidoxypropyltrimethoxysilane/diisopropylethylamine (100:30:1 v/v/v). The reaction took place at 80 °C with stirring over 24 h under an argon atmosphere. The substrates were then collected, successively washed with two 50 mL portions of each of methanol, dichloromethane, diethyl ether, and then dried and stored under vacuum and over P2O5 at room temperature until required. Surface Modification of Solid Substrates: Linkage of DMT-HEG onto GOPS-Functionalized Substrates. DMTHEG was synthesized as outlined previously.25 DMT-HEG (700 mg of DMT-HEG/100 mg of CPG) that had been dried under vacuum and over P2O5 (>72 h) was dissolved in 20 mL of anhydrous pyridine. An excess of NaH (10 equiv) that had been thoroughly washed with dry hexane was then introduced to the mixture. The subsequent reaction was permitted to proceed with stirring for 1 h at room temperature under an argon atmosphere. The reaction mixture was filtered through a sintered glass frit under a positive pressure of argon into a vessel containing the GOPS-functionalized substrates. GOPS-functionalized substrates were separated into three batches, containing both optical fibers and CPG. The three batches then underwent the DMTHEG coupling reaction, which was permitted to proceed under a positive pressure of argon at room temperature with gentle agitation on an oscillating platform stirrer for durations of 1, 4, and 12 h, respectively. Following the coupling reaction, the substrates were quickly recovered and washed with successive 150 mL portions of methanol, water, methanol, and diethyl ether to quench the coupling reaction and remove any reactants that were nonspecifically adsorbed. The substrates functionalized with (18) Kern, W.; Puotinen, D. A. RCA Rev. 1970, 6, 187. (19) Pon, R. T. In Methods in Molecular Biology: Protocols for Oligonucleotides and Analogues; Agrawal, S., Ed.; Humana Press Inc.: Totowa, NJ, 1993; 465. (20) Sojka, B.; Piunno, P. A. E.; Wust, C. C.; Krull, U. J. Anal. Chim. Acta 1999, 395, 273. (21) Uddin, A. H.; Piunno, P. A. E.; Hudson, R. H. E.; Damha, M. J.; Krull, U. J. Nucleic Acids Res. 1997, 25, 4139. (22) Chalikian, T. V.; Vo¨lker, J.; Plum, G. E.; Breslauer, K. J. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 7853. (23) Maskos, U.; Southern, E. M. Nucleic Acids Res. 1992, 20, 1679. (24) Piunno, P. A. E.; Watterson, J. H.; Wust, C. C.; Krull, U. J. Anal. Chim. Acta 1999, 400, 73. (25) Ohchepinov, M. S.; Case-Green, S. C.; Southern, E. M. Nucleic Acids Res. 1997, 25 (6), 1155.

4986

Langmuir, Vol. 16, No. 11, 2000

DMT-protected HEG were dried under vacuum and over P2O5 and were maintained under these conditions until further required. Surface Modification of Solid Substrates: Capping of Unreacted Silanol and Hydroxyl Functionalities with Chlorotrimethylsilane (TMS-Cl). Unreacted silanol and hydroxyl functionalities on the surface of the solid substrates where undesired oligonucleotide synthesis could occur were capped prior to oligonucleotide synthesis using TMS-Cl according to the method of Pon et al.19 The dried substrates were suspended in a solution of 1:10 (v/v) chlorotrimethylsilane/pyridine for 16 h under an argon atmosphere at room temperature. The substrates were subsequently recovered and washed with three successive 20 mL portions of pyridine, methanol, and diethyl ether and were then stored under vacuum and over P2O5 at room temperature until required. Solid Phase Phosphoramidite Synthesis of Oligonucleotides. All solid phase oligonucleotide synthesis was done using a PE-ABI 391-EP DNA synthesizer (Perkin-Elmer Applied Biosystems, Foster City, CA). The preprogrammed synthesis cycles employed for oligonucleotide assembly were modified to adjust the reagent delivery times in order to ensure that the synthesis columns used were completely filled. The column used for oligonucleotide synthesis onto optical fiber segments was a custom-manufactured Teflon synthesis column (6 mm i.d. × 50 mm) capable of holding 8 fibers, circularly distributed in a manner which ensured no interfiber contact. The fibers were secured by means of insertion into cylindrical bores (400 µm i.d. × 2 mm deep) machined into one of the end caps. All end caps were secured onto the column bodies with aluminum crimp seals. The columns used for oligonucleotide synthesis onto DMT-HEG-GOPS functionalized CPG were custom manufactured Teflon columns (8 mm i.d × 10 mm). Teflon end filters (0.22 µm pore size, PE-ABI) were used to retain the glass beads within the column. Synthesis of oligonucleotides for use as complementary material for immobilized DNA was carried out on nucleoside-functionalized LCAA-CPG substrates prepacked in polyethylene columns as supplied by the manufacturer. Detritylation was done using 3% trichloroacetic acid in dichloromethane or 2% dichloroacetic acid in dichloromethane. Activation of phosphoramidites for synthesis onto substrates was achieved with 0.5 M tetrazole in acetonitrile (LCAA-CPG substrates) or ethylthiotetrazole (DMT-HEG-CPG substrates). Reagents for acetylation of unreacted hydroxyl functionalities were prepared as follows: cap A, 10% acetic anhydride and 10% collidine in THF; and cap B, 16% Nmethylimidazole in THF (w/v). Oxidation was done with a solution of iodine, 0.1 M, in THF/pyridine/water (25:20:2, v/v/v). Prior to oligonucleotide synthesis, the derivatized solid supports were treated with the acetylating reagents by completing the capping portion of a standard synthesis cycle, to ensure blocking of any remaining hydroxyl functionalities. Phosphoramidite reagents were dissolved in dry, freshly distilled acetonitrile to a concentration of 0.1 M. Polythymidylic acid icosanucleotides (dT20) were assembled onto all of the optical fiber and CPG substrates functionalized with DMT-HEG linker molecules. Determination of the density of surface coverage of CPG substrates with covalently immobilized oligonucleotide-HEG conjugates was done by anion-exchange HPLC following methods developed in our research group that have been reported elsewhere.20 Icosanucleotides labeled at the 5′-terminus with a fluorescein moiety were used as complementary material to hybridize with immobilized dT20 sequences. The 5′-fluorescein labeled oligonucleotides were prepared by use of a fluorescein phosphoramidite synthon (Dalton) and standard protocols for oligonucleotide preparation.21 Additionally, unlabeled complementary icosanucleotides were prepared by standard protocols for use in studies of hybridization in bulk solution. Instrumentation for Studies of Immobilized Nucleic Acid Membranes. Fluorescence-based studies of nucleic acid hybridization at the surface of optical fibers were carried out via an automated spectrofluorimeter instrument developed in our research group that has been described previously.24 Laser radiation (488 nm) from a Coherent Innova 70 CW argon ion laser (Coherent Laser Products, Palo Alto, CA) was coupled into a sensing fiber by first guiding the beam of source radiation such that it was incident upon the surface of a dichroic mirror (505

Watterson et al. nm cutoff, Omega Optical, Battleboro, VT) oriented at 45° to the incident beam. The sensing fiber was secured via a Teflon holder with a waterproof compression-type seal within a stainless steel hybridization cell with small volume (1 mm i.d. × 50 mm) that permitted a solution volume of 137 µL to be exposed to the sensing fiber. Fluorescence emission from the sensing fiber with wavelength greater than 505 nm was then directed back through the dichroic mirror into a Bentham M300 monochromator (f/# ) 4.2, 2.5 nm bandwidth, distributed by Optikon Corporation Limited, Waterloo, ON). Fluorescence emission exiting the emission monochromator was detected by a side-on photomultiplier tube (106 A/W responsivity, model 77348, Oriel Corp.) operated at a potential of 500 V dc (PMT power supply model 5502, Products for Research, Danvers, MA). The temperature of the solution within the flow cell was determined by use of a glass-encapsulated bead thermistor (Fenwal Electronics Inc., distributed by Electrosonic Inc., Toronto, ON) embedded within the stainless steel block at a distance not more than 1 mm from the internal wall of the solution compartment surrounding the sensing fiber. The temperature of the flow cell was regulated by use of a Pelletier temperature control accessory (model 89090A, Hewlett-Packard Corp., Mississauga, ON). Analyte sampling and delivery was done using an automated sampler and pump system as previously described.24 Acquisition of Thermal Denaturation Profiles from Nucleic Acid Membranes Immobilized onto Fused Silica Optical Fibers. All thermal denaturation profiles for hybridization occurring at the surface of the optical fiber sensors were acquired by monitoring the intensity of fluorescence emission at 542 nm over the temperature range of ca. 20-80 °C using a temperature ramp rate of 0.3 °C min-1 to ensure that equilibrium conditions were satisfied. All sensors were cleaned by sonication in ethanol in a 40 W bath sonicator for 90 min to remove adsorbed impurities from the sensor surface prior to analysis. Thermal denaturation profiles were obtained for optical sensors that were exposed to mixtures of dA20 and dA20-5′-fluorescein in a 100:1 molar ratio, and d(A9GA10) and d(A9GA10)-5′-fluorescein (100:1 molar ratio) in various dilutions of a stock phosphate buffered saline (PBS) hybridization buffer (1.0 M NaCl, 50mM PO4-n, pH 7.0). Dilutions of the stock buffer by factors of 1.0, 0.5, and 0.1 were used for ionic strength studies. All analyses were done in triplicate for each ionic strength of PBS buffer investigated, and the standard deviation in Tm values was less than 1.6 °C for each sample set. Removal of complementary oligonucleotide associated with the sensor surface from previous analyses was done prior to each subsequent experiment by flushing 15 mL of 80 °C water through the flow cell (3 mL min-1, 5 min) and by flushing 1 mL of 90% formamide in TE buffer (10 mM Tris-HCl, 5 mM EDTA, pH ) 8.3) through the flow cell. Acquisition of Thermal Denaturation Profiles from Nucleic Acid Hybrids in Bulk Solution. Studies of nucleic acid hybridization in solution were carried out by measuring the absorbance at 260 nm of solutions containing oligonucleotides. All absorbance measurements were made with a Hewlett-Packard 8452A spectrophotometer interfaced to a PC by means of an HPIB interface (HP82335, Hewlett-Packard). Sample cuvettes were contained within a water-jacket cell holder built in-house with the capacity to hold four sample cells. The cell holder was situated in a track with an embedded, motor-driven, variable-speed stir magnet. The sample to be analyzed was moved into the optical path by actuation of a stepper motor modified such that it could be controlled through a standard PC parallel port. The temperature of the cell holder was controlled by means of water delivery from a temperature-controlled water bath (LAUDA RM20 recirculating water bath) modified in-house such that the temperature of the circulating water could be controlled via a standard PC parallel port. The temperature of the cell holder was measured by means of a thermistor (Fenwal Electronics Inc., distributed by Electrosonic Inc., Toronto, ON) embedded within the cell wall, not more than 3 mm from the circulating water. Temperature values were calculated by use of an empirical calibration equation that provided an accuracy in reported temperature values to within 0.1 °C. Thermal denaturation profiles for hybridization occurring in bulk solution were acquired over the temperature range of ca. 20-80 °C using a temperature ramp rate of 0.1-0.2 °C min-1 in


Langmuir, Vol. 16, No. 11, 2000 4987

Table 1. Tm °C Values Obtained for dT20 Hybridized with Various Oligonucleotides, in Hybridization Buffers of Various Ionic Strengths thermal denaturation temp, Tm (°C), for dT20 hybridized with various oligonucleotides (total [dsDNA] ) 0.62 µM, equimolar amounts of ssDNA) dA20 [NaCl] (M) 1.0 0.75 0.50 0.30 0.1 ∂Tm/∂ log [Na+] (°C) R2

57.6 ( 0.4 55.7 ( 0.5 53.9 ( 0.5 50.6 ( 0.6 43.3 ( 0.6 14.2 ( 0.3 0.998

d(A9GA10) 53.8 ( 0.5 51.7 ( 0.4 48.3 ( 0.4 44.5 ( 0.5 37.4 ( 0.6 15.6 ( 0.7 0.994

an effort to ensure that equilibrium conditions were satisfied. Thermal denaturation profiles were obtained for dT20 hybridized in equimolar amounts with each of the following oligonucleotides: dA20, d(A9GA10), d(A9G2A9), d(A18G2), d(G2A16G2), and d(G5A20G5). Analyses were done in various dilutions of a stock phosphate buffered saline (PBS) hybridization buffer (1.0 M NaCl, 50mM PO4-n, pH 7.0). Dilutions of the stock buffer by factors of 1.0, 0.75, 0.5, 0.3, and 0.1 were used for ionic strength studies. All analyses were done at least twice for each ionic strength of PBS buffer investigated.

d(A9G2A9) 48.7 ( 0.6 46.6 ( 0.3 44.2 ( 0.3 41.0 ( 0.4 33.2 ( 0.5 15.4 ( 0.3 0.999

d(A18G2) 56.9 ( 0.4 55.1 ( 0.5 52.3 ( 0.4 49.8 ( 0.4 43.0 ( 0.5 13.8 ( 0.3 0.999

53.5 ( 0.7 50.8 ( 0.7 49.2 ( 0.5 44.9 ( 0.6 37.0 ( 0.6 16.3 ( 0.7 0.995

transition. While not a strictly accurate representation of the thermal denaturation of a 20-mer and the fact that partial hybridization is possible in a dA20-dT20 system, this commonly used approach is applied for consistency in relative comparisons of solution-based and surfaceimmobilized hybridization. For solutions containing equimolar amounts of complementary oligonucleotides, the van’t Hoff enthalpy change at Tm is computed from the normalized thermal denaturation data as described by Breslauer:16

Results and Discussion Studies of Nucleic Acid Hybridization in Bulk Solution. Thermal denaturation profiles were obtained for oligonucleotides hybridized in bulk solution to determine some of the trends in the thermodynamics of hybridization. These data are important as control experiments because the ssDNA was produced by automated synthesis and will therefore have similar heterogeneity in comparison to immobilized ssDNA to partial synthesis failure during each step in the synthesis procedure. Initial experiments consisted of an examination of the relationship between the observed thermal denaturation temperature, Tm, and the ionic strength of the hybridization solution. In these experiments dT20 (0.62 µM) was hybridized with one of the following oligonucleotides in a 1:1 molar ratio: dA20, d(A9GA10), d(A9G2A9), d(A18G2), d(G2A16G2), or d(G5A20G5). Hybridization was carried out in a solution of PBS (1 M NaCl, 50 mM NaH2PO4, 50 mM Na2HPO4) buffer diluted by a factor of 1.0, 0.75, 0.5, 0.3, or 0.1. Experimental data was analyzed on the basis of the two-state hybridization model described by Breslauer.16 The values of Tm obtained as a function of hybridization buffer ionic strength are tabulated in Table 1 for all oligonucleotide hybrids used. The respective rates of change of Tm with respect to log [Na+] are also shown in Table 1, along with the correlation coefficients (R2) for these relations. These correlation coefficients show that the relation between Tm and log [Na+] displays very good linearity over the entire ionic strength regime used. The results confirm that the presence of base-pair mismatches has the potential to reduce the observed Tm value of the duplex. Furthermore, the deviation in Tm for a duplex that contains base-pair mismatches, in comparison to that of the fully complementary duplex is a function of the ionic strength of the hybridization solution, the number of base-pair mismatches, and their positions within the duplex.12 This is consistent with trends in oligonucleotide hybridization trends that have been previously reported.17 Differences in the hybridization thermodynamics between fully complementary hybrids and those that contain base-pair mismatches may be examined by computing the van’t Hoff enthalpy change accompanying the denaturation transition. The van’t Hoff enthalpy change is computed assuming that denaturation is a two-state

d(G2A16G2)

( )

∆HVH,Tm ) -6RTm

∂fss ∂T

T)Tm

(1)

Values of ∆HVH as a function of temperature may be obtained by computing values for a given duplex in hybridization buffer at various ionic strengths. Recently, Breslauer22 reported that the enthalpy change accompanying denaturation was in fact a function of temperature as a result of a small change in the heat capacity of the system that was associated with denaturation. This finding is contrary to assumptions made hitherto in studies of oligonucleotide hybridization thermodynamics.22 It is therefore possible to use values of ∆HVH obtained at Tm in hybridization buffers of different ionic strengths to compute values of ∆H° at a standard reference temperature, to establish a basis of comparison for the relative stability of two different sequences. In general, the enthalpy change for a given process is a function of temperature according to the following relation:

∆H°T ) ∆H°Tref +

∫TT ∆Cp dT ref

(2)

The value of ∆Cp may be obtained by computing the slope of a plot of ∆HVH versus Tm from denaturation experiments in hybridization buffers of different ionic strengths. The values of ∆HVH at Tm and ∆H° corrected to 40 °C for dA20: dT20 and d(A9GA10):dT20 in hybridization buffers of various ionic strengths are shown in Table 2. These values correspond to an experimentally determined value of ∆Cp of 112 ( 3 cal deg-1 molbp for dA20:dT20, which is in good agreement with the value for the polymeric duplex poly(dA):poly(dT) presented by Breslauer23 (101.7 ( 24 cal deg-1 molbp). Studies of Immobilized Nucleic Acid Hybridization at an Interface. Thermal denaturation profiles were obtained for oligonucleotides that were covalently immobilized to the surface of fused silica optical fibers. The results were used to examine differences between trends observed for hybridization experiments that were done in bulk solution and those of DNA hybridization at an interface, as well as to determine the effects of immobilization density on the thermal denaturation profiles of immobilized oligonucleotides. Since many nucleic acid

4988


Watterson et al.

Table 2. van’t Hoff Enthalpy Changes at Tm and Corrected to 40 °C for 0.62 µM Solutions of dA20:dT20 and d(A9GA10):dT20 in Hybridization Buffers of Various Ionic Strengths dA20:dT20 [NaCl] (M)

Tm (°C)

1.0 0.5 0.3

57.6 ( 0.4 53.9 ( 0.5 50.6 ( 0.3

d(A9GA10):dT20

∆HVH,Tm (kcal/mol)

∆H° (40 °C) (kcal/mol)

168 ( 4 162 ( 4 153 ( 3 mean

128 131 138 129 ( 2

Tm (°C) 53.8 ( 0.4 48.3 ( 0.4 44.5 ( 0.5



137 ( 5 138 ( 5 118 ( 5 mean

105 114 106 108 ( 5

Table 3. Density of Immobilization of dT20-HEG Conjugate onto GOPS-Functionalized Substrates as Determined by Anion-Exchange High-Performance Liquid Chromatography sample density low medium high

reactn duration total surface molecules av radius/ (DMT-HEG- area of CPG dT20-HEG molecule substrate) (h) used (Å2) immobilized (Å) 1 4 12

2.62 × 1019 2.62 × 1019 4.12 × 1019

2.41 × 1014 1.15 × 1015 1.90 × 1016

186.2 85.3 26.3

biosensor and biochip schemes involve hybridization of oligonucleotides immobilized to a solid surface, it is of obvious importance to establish trends in the hybridization thermodynamics for such systems in order to address issues of sensitivity and selectivity. Immobilization Density Considerations. The immobilization of polythymidylic icosanucleotides (dT20) onto the surface of fused silica optical fiber substrates was achieved by means of a modification to the method of Maskos and Southern.23 The fused silica optical fiber substrates were first functionalized with glycidoxypropyltrimethoxysilane (GOPS) as described above. Hexaethylene glycol (HEG), protected on one terminus with dimethoxytrityl (DMT) in order to ensure single-site reactivity and to minimize the risk of formation of closedring structures, was then covalently attached to the epoxysilane layer. The modified optical fiber substrates were then subjected to standard β-cyanoethyl-phosphoramidite oligonucleotide synthesis protocols to prepare by stepwise synthesis the dT20 oligonucleotides on the surface of the substrates. One goal of this study was to determine if the density of immobilized DNA affects the thermodynamics of hybridization and, therefore, the selectivity of a hybridization assay at the surface of an optical fiber biosensor. The density of oligonucleotide immobilization was controlled by means of controlling the reaction time of DMTHEG conjugates with the GOPS-functionalized substrates. Three different immobilized densities were obtained by permitting the DMT-HEG coupling reaction to proceed for 1, 4, and 12 h, respectively. To characterize the density of immobilization, oligonucleotide synthesis was carried out as described above on GOPS-functionalized controlledpore glass (CPG), which has a well-defined surface area, in tandem with the oligonucleotide synthesis on the optical fiber substrates. The oligonucleotide-HEG conjugates were then cleaved from the surface of the CPG by means of exposure to concentrated ammonium hydroxide for approximately 3 h, lyophilized, and redissolved in water. The immobilization densities as well as the quality of automated synthesis of all immobilized oligonucleotide samples were subsequently analyzed by anion-exchange HPLC. Quantitation of the cleaved HEG-dT20 conjugates was achieved by coinjection with a known quantity of dT20. The results of the HPLC analysis are shown in Table 3, and the chromatograms are shown in Figure 1. The peaks corresponding to HEG-dT20 were determined to have retention times of 25.5-26 min, according to previous

Figure 1. Normalized AE-HPLC chromatograms obtained in analysis of packing density of dT20-HEG immobilized to surface of CPG, after cleavage and recovery from concentrated ammonium hydroxide. Chromatograms correspond to samples with different reaction time between GOPS-functionalized CPG and DMT-HEG: coupling time: (a) 1 h, (b) 4 h, (c) 12 h.

studies reported by our research group.20 Distributions of peaks in the region of 26 min were not attributed to species corresponding to failed nucleotide coupling in the automated synthesis protocol, but were due to the formation of a HEG-linker membrane in which some of the HEG molecules were inadvertently coupled onto one another to provide a distribution of linker lengths. This was corroborated by the fact that nucleotide coupling efficiencies were always greater than 99%. The data show that the three densities used in thermal denaturation experiments were representative of three different physical environments for the immobilized oligonucleotides. The low-density sample consisted of immobilized dT20-HEG conjugates separated by approximately 372.4 Å between adjacent strands, assuming uniform oligonucleotide distribution. Since the length of the dT20-HEG conjugate is ca. 100 Å, the low-density sample then represents the system wherein there is, on


average, very little chance of interactions between neighboring strands that may affect hybridization. The mediumdensity sample consisted of immobilized dT20-HEG conjugates separated by approximately 170.6 Å between adjacent strands, which may permit the onset of some interaction between neighboring strands. Finally, the high-density sample consisted of immobilized dT20-HEG conjugates separated by approximately 52.6 Å between adjacent strands. This close packing is much more likely to facilitate interactions between neighboring strands than the lower packing densities. Thermal Denaturation of Surface Immobilized Oligonucleotide Hybrids. Optical fibers that were prepared by synthesizing dT20-HEG conjugates onto the surface of functionalized fused silica were subjected to hybridization and thermal denaturation experiments. Complementary oligonucleotide solutions contained mixtures of unlabeled dA20 and dA20-5′-fluorescein, together in a 100:1 molar ratio, with a total oligonucleotide concentration of 10-7 M. It was assumed that the fluorescein label would not seriously impede the hybridization process, and a control thermal denaturation experiment was conducted in bulk solution using dA205′-fluorescein and dT20 as the complementary oligonucleotides (0.5xPBS, 0.62 µM dsDNA, equimolar in each strand). The observed Tm was 55.1 ( 0.8 °C, which is in reasonable agreement with that observed in an analogous experiment using unlabeled dA20, described above (53.9 ( 0.5 °C). Similarly, studies of hybridization thermodynamics of sequences containing a centrally located SBPM were done using analogous mixtures of unlabeled and labeled d(A9GA10) in the same molar ratio and with the same total oligonucleotide concentration. Excitation radiation was delivered to the surface of an optical fiber by means of coupling a beam from an argon ion laser into the optical fiber. The fluorescent emission was coupled back into the optical fiber and collected at a wavelength of 542 nm. The temperature was ramped in these experiments over the range 25-100 °C at a rate of 0.3 °C min-1. Complementary oligonucleotides were introduced in hybridization buffers of various ionic strengths (0.1, 0.3, 0.5, or 1 M NaCl) to establish the trends in interfacial hybridization thermodynamics as they relate to the ionic strength of the hybridization solution. An example of the raw data obtained from the thermal denaturation profiles measured at the surface of the optical fiber with medium packing density of immobilized dT20 is shown in Figure 2. Again, the data was analyzed on the basis of assumption that the denaturation took place as a two-state transition. The upper and lower baselines were used to extrapolate the thermal fluorescence decay. It was assumed that full hybridization occurred and that the oligonucleotides were fully denatured in the high-temperature regime. Monoexponential decay profiles were fitted to the baseline data in the case of fluorescence measurements. This type of profile was found to accurately describe the thermal fluorescence decay that was obtained when a fused silica fiber treated with only trimethylchlorosilane (TMS-Cl) was exposed to a solution of dA20 and dA20-5′-fluorescein and was subjected to the same temperature ramping conditions (data not shown). Conversion of the raw data to a normalized thermal denaturation profile consisting of the normalized fraction of ssDNA present as a function of temperature was then achieved by treatment analagous to that used for the raw absorbance data, with consideration given to the inequality of molar amounts of each strand. All interfacial thermal denaturation profiles were then reported such that the values of the fraction of ssDNA were normalized to fall

Langmuir, Vol. 16, No. 11, 2000 4989

Figure 2. Normalized thermal denaturation plots for dA20/ dA20-fluorescein for the optical biosensor with low immobilized ssDNA packing density in PBS buffers of various ionic strengths.

Figure 3. Thermal denaturation profile for 10-7 M dA20/dA20fluorescein for the optical biosensor with medium immobilized ssDNA packing density in 1.0 × PBS buffer.

between values of 0 and 1, to simplify visual comparison. All thermodynamic data were computed on the basis of actual, unnormalized data, so this normalization did not affect reported results. The normalized thermal denaturation profiles obtained using the optical fiber biosensor with low oligonucleotide packing density in hybridization buffers of different ionic strengths and using dA20/dA20-fluorescein as the complementary material are shown in Figure 3. The Tm data observed as a function of ionic strength for the low, medium, and high packing densities and using dA20/dA20fluorescein as the complementary material are shown in Table 4. The data in Table 4 illustrate the effect of packing

4990


Watterson et al.

Table 4. Observed Tm (°C) Values for Optical Fiber Biosensors with Low, Medium, and High Oligonucleotide Packing Densities Using Hybridization Buffers of Various Ionic Strengths and dA20/dA20-5′-fluorescein as the Complementary Material Tm (°C)

[NaCl] (M) 0.1 0.5 1.0 ∂Tm/∂ log [Na+] (°C)

low packing density

medium packing density

high packing density

39.5 ( 0.2 50.7 ( 0.2 54.9 ( 0.2 15.5 ( 0.5

41.6 ( 0.2 48.0 ( 0.2 53.1 ( 0.2 11 ( 2

32.3 ( 0.2 43.5 ( 0.2 46.4 ( 0.2 14 ( 1

density on the thermodynamics of hybridization. The high packing density facilitated some destabilization of the hybridized immobilized oligonucleotides as evidenced by the Tm values which were consistently lower than those observed with the low packing density and medium packing density optical fiber biosensors. Additionally, the sensitivity of Tm to salt concentration in the hybridization buffer appeared to be fairly consistent with observations made in bulk solution, and the three values reported in Table 4 agree within experimental uncertainty at the 95% confidence level. This supports the notion that there is no significant difference in the ion environments within the nucleic acid membranes brought about as a function of oligonucleotide packing density, as predicted by a theoretical model developed in our research group.24 It may be that the differences in Tm observed with the optical fiber biosensor with high oligonucleotide packing density relative to those with the low and medium packing densities are a result of greater interaction between neighboring immobilized strands as well as the interaction of strands with the solid surface, whereby such interactions interfere with and reduce the stability of the WatsonCrick interactions. Interactions between immobilized strands may also reduce the number of immobilized oligonucleotides that are available for hybridization, similar to what has been reported by Southern.25 Nearestneighbor interactions between immobilized strands and between the immobilized strands and the solid substrate surface may reduce the validity of the assumption of a two-state transition, as is suggested by the broadening of the interfacial denaturation transition relative to that in bulk solution (data not shown). Using a two-state model in the analysis of hybridization both in bulk solution and in interfacial environments allows a comparison of the relative validity of the model for each system to be made. The broadening of the denaturation transitions observed for interfacial hybrids may suggest the presence of intermediate complex structures which exist throughout the denaturation process that may in turn affect the ability of thermal methods to be used to control hybridization stringency. To establish trends in the hybridization energetics which govern selectivity, thermal denaturation experiments identical to those described above were done using the low, medium, and high packing density optical fiber biosensors and d(A9GA10)/d(A9GA10)-fluorescein as the complementary material. The observed Tm values in those experiments are listed in Table 5. It should be noted that thermal denaturation data for the optical fiber biosensor with high oligonucleotide packing density displayed significantly greater noise than did the data from the other fluorescence experiments. Examination of the data in Tables 4 and 5 shows that, for the optical fibers with low and medium oligonucleotide packing density, the deviations in Tm caused by the

Figure 4. Deviation in observed Tmvalues, ∆Tm (°C), as a result of the presence of a centrally located SBPM for the low, medium, and high immobilized oligonucleotide packing densities, in hybridization buffers of various ionic strengths. Table 5. Observed Tm (°C) values for Optical Fiber Biosensors with Low, Medium, and High Oligonucleotide Packing Densities Using Hybridization Buffers of Various Ionic Strengths and d(A9GA10)/ d(A9GA10)-fluorescein as the Complementary Material Tm (°C)

[NaCl] (M) 0.3 0.5 1.0 ∂Tm/∂ log [Na+] (°C)

low packing density


high packing density

39.2 ( 0.6 42.4 ( 0.5 48.5 ( 0.5 18 ( 2

39.2 ( 0.6 42.0 ( 0.5 45.9 ( 0.5 12.7 ( 0.1

31.1 (_1.6 33.5 ( 1.0 36.5 ( 1.1 10.3 ( 0.2

presence of a centrally located SBPM were larger when hybridization occurred in solutions of lower ionic strength, relative to those observed in experiments done in hybridization buffers of higher ionic strength. This observation is consistent with observations made in thermal denaturation experiments conducted in bulk solution, as shown in Table 1. The preliminary results indicate that the opposite trend was observed with the biosensor with high oligonucleotide packing density. It may be that the higher packing density of immobilized DNA permits greater interaction between neighboring strands under conditions of increased ionic strength within the hybridization solution and the immobilized nucleic acid layer. This could result in greater destabilization of the WatsonCrick bonding within the hybrids and would lead to greater deviations in the observed Tm for solutions of higher ionic strength. The deviations in observed Tm values between thermal denaturation experiments using dA20/dA20-5′fluorescein as the complementary oligonucleotide and those using d(A9GA10)/d(A9GA10)-5′-fluorescein as the complementary oligonucleotide are shown graphically in Figure 4. Enthalpy Considerations. A comparison of the data shown in Tables 4 and 5 with that shown in Table 1 shows that deviations in Tm caused by the presence of a centrally located SBPM were significantly larger for experiments involving immobilized dsDNA relative to those observed for dsDNA floating freely in bulk solution. This observation is significant since it suggests that the thermodynamic selectivity of a hybridization assay using immobilized DNA


Langmuir, Vol. 16, No. 11, 2000 4991

Table 6. van’t Hoff and Standard Enthalpy Changes for Denaturation of Immobilized Oligonucleotides with Different Packing Densities and Ionic Strengths low packing density ∆HVH,Tm (kcal/mol)

[NaCl] (M)





high packing density ∆HVH,Tm (kcal/mol)


1.0 0.5 0.1

(a) Using dA20/dA20-fluorescein (10-7 M) as the Complementary Oligonucleotide 34 ( 2 41.5 30 ( 3 44.9 34.8 ( 3 37 ( 2 42.4 35 ( 3 44.1 35.4 ( 3 42 ( 3 41.7 43 ( 3 44.8 65.6 ( 3 mean 42 ( 1 mean 45 ( 1 mean

48.0 36.8 49.4 45 ( 7

1.0 0.5 0.3

(b) Using d(A9GA10)/d(A9GA10)-fluorescein (10-7 M) as the Complementary Oligonucleotide 50 ( 1 37.1 36 ( 4 39.9 38 ( 5 40 ( 1 36.4 34 ( 2 35.4 39 ( 5 36 ( 2 37.2 41 ( 4 40.5 39 ( 2 mean 37 ( 1 mean 39 ( 3 mean

37.3 37.8 37.3 37 ( 1

may be significantly better than what may have otherwise been predicted by thermal denaturation experiments conducted in bulk solution. This enhancement of the deviation in the Tm values as a result of the presence of a centrally located SBPM suggests that both hydrogenbonding energetics and entropic effects associated with hybridization may be quite different in the interfacial environment than they may be in bulk solution. To examine the energetics of interfacial hybridization, the van’t Hoff enthalpy changes and temperature-corrected standard enthalpy changes were computed for each of the denaturation experiments conducted here on the basis of the method developed in our research group.24 This model applies to denaturation occurring within a membrane of immobilized nucleic acids, with the complementary DNA freely able to move into and out of the membrane. The model assumes no interaction between neighboring strands, and that the denaturation is a two-state transition. With these assumptions, any experimental data which appears to show deviations from a two-state model become illustrative of extrahybrid interactions between nearest neighbors or the between the immobilized strands and the solid substrate surface. The van’t Hoff enthalpy change is then given by the equation

[( ) ( (

∆HVH,Tm ) -

) )] ( )

1 - fssmin 1 + + 1 - fss fss - fssmin fss - fssmin -1 AT ∂fss -1+ RTm BT 1 - fssmin ∂T

(3)

T)Tm

where fss is the total fraction of ssDNA present at any given time, fssmin is the minimum fss possible in a system with a nonstoichiometric number of complementary strands, AT represents the total molar amount of complementary oligonucleotide, and BT represents the total molar amount of immobilized probe oligonucleotide. The value of fssmin is then computed by the following equation:

fssmin )

|BT - AT| AT + BT

(4)

The van’t Hoff enthalpy changes at Tm, and the standard enthalpy changes corrected to a temperature of 40 °C for the different complementary oligonucleotides, oligonucleotide packing densities, and hybridization buffer ionic strengths used in these experiments, are summarized in Table 6. Temperature corrections were made as described above according to the method of Breslauer.23 The reference temperature used for all such corrections was chosen on the basis that it is an operational temperature commonly used for hybridization assays conducted in our

research group, chosen in order to enhance selectivity and hybridization kinetics. The data in Table 6 show that the enthalpic change accompanying denaturation in an interfacial environment is significantly lower than that which is observed in experiments conducted in bulk solution, as also indicated by data in Table 2. This suggests that there may be significant differences in the extent of Watson-Crick bonding in an interfacial environment compared with that which occurs in bulk solution. There did not appear to be a relationship between the packing density of immobilized oligonucleotides and the reduction in the endothermicity of the denaturation. Thus, since the observed Tm values are still of a magnitude comparable to those determined for experiments done in bulk solution, it is likely that there is a significant difference in entropy changes accompanying hybridization and denaturation in an interfacial environment, relative to those observed in experiments done in bulk solution. These differences in the entropy changes accompanying hybridization or denaturation may be dependent upon the density of immobilized oligonucleotides, as this parameter may also be affected by the extent of interaction between neighboring strands. Computation of entropy changes accompanying hybridization or denaturation in an interfacial environment would require computation of equilibrium constants for the hybridization process, which in turn requires knowledge of the ionic strength within the nucleic acid membrane.24 Similar computations for processes occurring in bulk solution have been known to introduce significant error,16 and these computations for the case of immobilized nucleic acid systems will be left as future work. Any structural restriction or reduction in the strength of the Watson-Crick bonding within interfacial nucleic acid hybrid base pairs may help contribute to the deviations in Tm that have been observed. Salt present in the hybridization buffer may facilitate electrostatic interactions between the polyanionic phosphate backbone of the immobilized DNA and any charged functionalities present on the surface of the fused silica substrate. This interaction between immobilized strands and the surface of the solid substrate may restrict the changes in oligonucleotide secondary structure accompanying hybridization, which may alter the observed entropy change accompanying the hybridization or denaturation process. This interaction may also reduce the strength of WatsonCrick interactions, which may be responsible for the reduction in the observed enthalpy changes accompanying the hybridization and denaturation process. The data also suggest that the magnitude of the van’t Hoff enthalpy change accompanying denaturation in an interfacial environment does not display the same sen-

4992


sitivity to changes in ionic strength and, therefore, Tm, as was observed for experiments conducted in bulk solution. The sensitivities of ∆HVH,Tm to changes in Tm were a factor of 2-4 smaller for the transitions occurring at the interface of the optical fibers relative to those observed for the experiments conducted in bulk solution. The sensitivities of ∆HVH,Tm to changes in Tm were a factor of 2-4 smaller for the transitions occurring at the interface of the optical fibers relative to those observed for the experiments done in bulk solution, and they were usually opposite in sign. This suggests that the changes in heat capacity that accompany denaturation are not the same in an interfacial environment as in bulk solution. This further supports the notion that interfacial hybridization occurs in a physical environment that is significantly different from that of hybridization in bulk solution. Summary and Conclusions Thermal denaturation profiles were generated for immobilized short oligonucleotide hybrids as well as for hybrids formed in bulk solution in order to compare trends in the thermodynamics of hybridization in these two different environments. While it was observed that the Tm values obtained from immobilized nucleic acid hybrids displayed the same general trends with respect to changes of solution ionic strength and the presence of single basepair mismatches as observed from hybrids formed in bulk solution, some important differences in the thermodynamics of hybridization of the two systems do exist. First, the presence of a centrally located SBPM resulted in reductions in the observed Tm values that were significantly greater for the immobilized hybrids (∼6-10 °C) than those observed for hybrids formed in bulk solution (∼4-6 °C). Second, it was observed that there was a 2-4-

Watterson et al.

fold reduction in standard enthalpy change accompanying the denaturation event for hybrids formed in an interfacial environment in comparison to those formed in bulk solution. These observations suggest that the physical environment of hybrids formed at a solid interface is significantly different from that of hybrids formed in bulk solution. This notion is corroborated by the observation of changes in heat capacity accompanying the denaturation event at an interface that are not consistent with those observed in experiments done in bulk solution. These effects may be the result of interactions between the immobilized oligonucleotides and neighboring strands, as well as between the immobilized strands and the solid substrate surface. These interactions may interfere with the process of hybrid formation and conventional WatsonCrick base pairing. For analytical applications, these effects have practical implications for the selectivity and sensitivity of hybridization assays done using biosensors and biochips. Enhancement of the deviations in Tm as a result of centrally located SBPMs for systems of immobilized oligonucleotides allow for greater control of selectivity for hybridization assays. Unfortunately, interactions between neighboring strands and between immobilized strands and the surface of the solid substrate may affect the availability of the immobilized strands for hybridization. This effect may be dependent upon the density of immobilized oligonucleotides, which implies that there may be variations in sensitivity from sensor to sensor, depending on the ability to control the density of immobilization and the availability and chemical properties of any exposed substrate surface. LA991508M

Effects of Oligonucleotide Immobilization Density on Selectivity of

Recommend Documents