Mismatching base-pair dependence of the kinetics of DNA–DNA hybridizat
http://www.100md.com
《核酸研究医学期刊》
1 Max-Planck-Institute for Polymer Research, Ackermannweg 10, D-55128 Mainz, Germany and 2 Research Institute for Cell Engineering, National Institute of Advanced Industrial Science and Technology (AIST), 1-8-31 Midorigaoka, Ikeda, Osaka 563-8577, Japan
*To whom correspondence should be addressed. Tel: +49 6131 379160; Fax: +49 6131 379360; Email: knoll@mpip-mainz.mpg.de
ABSTRACT
Two single-stranded DNAs consisting of complementary base pairs except for one mismatching base pair (MM1) can form double-stranded DNA by molecular recognition. This type of duplex is not as stable as that formed by MM0. In order to add to a better understanding of the physical mechanism of the hybridization and dissociation processes at sensor (chip) surfaces, we studied the kinetics of the MM1 hybridization by surface plasmon fluorescence spectroscopy. Target DNA strands labelled with a fluorescent molecule Cy5 at the 5' end and hybridizing with the surface-attached probe DNA can be excited by the strong optical field of a surface plasmon resonance mode. The emitted fluorescence can be detected with high sensitivity. The affinity of a duplex was found to depend on the chemical nature, i.e. G–G, G–T etc., and on the position of the mismatching base pair along the 15mer duplex.
INTRODUCTION
The need for optical techniques sensitive enough to detect mismatching base pairs was greatly increased by the urgent request to identify point mutations associated with cancer and other diseases. Especially, technologies based on the DAN-chip platform (1–4) using hybridization strategies between a single-stranded probe DNA immobilized on the sensor surface and a complementary target strand binding from solution are required. Therefore not only DNA–DNA but also peptide nucleic acid (PNA)–DNA hybridization reactions have been studied by various groups as a very important research subject, applying a broad range of experimental techniques (5–9). The usual methods employed to observe such hybridization reactions are based on monitoring the mass deposition measured with the quartz crystal microbalance (QCM) (10–12) or recording the change in the optical thickness at the surface detected by surface plasmon spectroscopy (SPS) (13–16). Recently we introduced surface plasmon fluorescence spectroscopy (SPFS) as a technique that offers higher detection limits and higher sensitivity than the above mentioned methods (16,17). In order to evaluate the kinetics of DNA–DNA hybridization without intermolecular coulombic interactions between neighbouring binding sites, our strategy is to limit the probe DNA to a maximum density of 2.5 x 1012 cm–2, which was shown to be still near the limit of infinite dilution by binding the probes via their biotin end to a streptavidin monolayer. In such a dilute surface, hybridization for 15 bp between probe and target is difficult to observe by SPR quantitatively; however, a quantitative analysis is possible by the in situ recording of the kinetics of the association (hybridization) and dissociation process (18) by SPFS using target oligo-DNA chemically labelled with a fluorescent molecule. The kinetic parameters obtained by SPFS can be theoretically analyzed by a Langmuir adsorption model, i.e. the rate constants for hybridization and dissociation can be evaluated for any hybridization systems. This is a real advantage compared with other experimental techniques; hence we consider SPFS as one of the most powerful experimental methods even though the target DNA has to be modified with a fluorescent molecule. However, this label does not modify the measured kinetic parameters as we could show independently by recent measurements in which the chromophore was attached to the probe strand and unlabelled target oligonucleotides hybridized with virtually identical rate constants. Moreover, a recent novel label-free detection principle based on surface plasmon diffraction which has a higher sensitivity than classical SPR, and hence allows for the recording of oligonucleotide hybridization and dissociation (though with a weaker signal-to-noise ratio compared with SPFS) without a chromophore attached to either single strand, confirmed the negligible influence of the presence of the fluorophore in the measured kinetic rate constants and affinity constants.
So far, we have shown that mismatch discrimination is easily achieved with high sensitivity (19). However, details of the effect of mismatching (20–22) and matching (K. Tawa, D. Yao and W. Knoll, in preparation) base pairs were not investigated. Therefore more experimental studies concerned with the type and position of a mismatching base pair in oligo-DNA sequences were considered to be necessary in order to clarify the mechanism of hybridization. Here, we focus on how the position and the chemical nature (combination) of a mismatching base pair can contribute to the kinetics of hybridization between 15mer oligonucleotide strands.
MATERIALS AND METHODS
Oligo-DNA samples (purification by HPSF) were obtained from MWG-biotech AG. The sequence of the various base units is collected in Table 1. The target oligonucleotides composed of bases complementary to the probe DNA are called MM0. The target DNA with one mismatching base is called MM1. Each oligo-DNA in the MM1 series has 15 base units with different positions of the mismatching base pair and different chemical nature (combinations) of base mismatches. The details of SPFS were reported in previous papers (16–19, K. Tawa, D. Yao and W. Knoll, in preparation). Briefly, the set-up that we use is based on essential modules of a ‘normal’ surface plasmon spectrometer as shown in Figure 1A (17,18). A He–Ne laser operating at 632.8 nm passes an optical chopper (used also as the reference for the lock-in amplifier) and two polarizers for intensity and polarization control. By using a –2 goniometer, the light reflected at the substrate covered with a thin gold layer is collected through a lens and monitored by a photodiode. If the reflectivity is observed at a fixed angle of incidence, any change of the interfacial architecture (e.g. induced by adsorption or desorption processes) can be quantified by evaluating the resulting change of the reflected intensity (16). The intensity of surface plasmon excited fluorescence was also measured by a photomultiplier at a fixed angle of excitation for which the reflectivity was R = 0.30. The time-dependent fluorescence intensity observed by SPFS corresponds to the kinetics of hybridization and dissociation between probe and target DNAs (17).
Table 1. Sequences of the oligo-DNA samples employed
Figure 1. (A) SPFS set-up. (B) Sample layers used for the DNA–DNA hybridization studies.
As has been discussed in a previous paper (18), the optical field which is strongly enhanced at the surface plasmon (PSP) resonance angle can be used to excite fluorescence molecules attached to the surface-bound analyte. The angular characteristic of the fluorescence intensity then follows the angular dependence of the optical field at the interface. While in the case of total internal reflection only an intensity enhancement of a factor of 4 can be achieved at the critical angle of total internal reflection, enhancement factors of up to about 20 can be achieved for gold. Fluorescent molecules that are within the evanescent field of the PSP, which decays normal to the interface with a penetration depth of Lz = 150 nm, can be excited by this surface plasmon mode. The emission is monitored by a photomultiplier (after passing through a set of interference filters) mounted to the goniometer part that rotates with , as does the coupling prism (see Fig. 1).
In order to achieve an optimized fluorescence signal and not to lose too much intensity by quenching to the thin gold layer (F?rster transfer), the chromophores have to be sufficiently separated from the gold surface. The relevant length scale is given by the F?rster radius which amounts to 5–7 nm. The molecular architecture of the interfacial layer, including the oligo-DNA layer prepared on the substrate, was as in our previous studies (as shown in Fig. 1B) (16–19). We achieve this by building the biotin/streptavidin architecture that ensures the required separation of the chromophores attached to the target DNA strands. First, a mixed self-assembled monolayer (SAM) of a biotinylated thiol derivative and a short-chain OH-terminated thiol used as a diluent at a mixing ratio of 1:9 was prepared on the gold substrates in order to optimize the formation of a streptavidin monolayer via the specific recognition to the biotin moieties. Then, the next layer of oligonucleotide catcher probes with a biotin group at their 5' end are organized in a rather ordered way at that interface. The density of probe DNA thus obtained corresponds to 2.5 x 1012 cm–2. The probe and target sequences used in this study are summarized in Table 1. All target sequences are labelled by a fluorophore called Cy5 at their 5' end. The sample cell employed is made from quartz glass and contains in- and outlets for the various solutions. It is sealed via a Viton O-ring against the gold-coated high refractive glass slide which is index-matched to the prism.
All solutions used in this study were prepared using phosphate-buffered saline (PBS) tablets (Aldrich-Sigma) with an Na+ concentration of 0.137 M. All measurements were performed at room temperature (22–23°C). The concentration of the target DNA solution was 0.25 μM for the measurement of the hybridization kinetics, and the solutions were applied via injection of 200–300 μl with a syringe. Once a stable surface coverage of bound target was reached, the dissociation process was initiated by rinsing pure PBS solution with a peristaltic pump through the cell.
RESULTS AND DISCUSSION
Hybridization kinetics: mismatching base-pair position dependence
Major factors that were reported to contribute to the interfacial binding process are the ionic strength and the pH in the buffer solution (23), the temperature (23,24), the sequence length and the chemical nature (combination) of base units (24). The comparison of the kinetics of the P2–T1(TG15), P2–T1(TG13), P2–T1(TG09) and P2–T1(TG02) probe–target pairs should be helpful for the understanding of the mismatching base-pair position dependence of the hybridization process. The corresponding kinetic curves are shown in Figure 2. As it turned out, a simple Langmuir model describing the association as well as the dissociation step by a single rate constant kon and koff, respectively, could not fit the data appropriately. Therefore an extended Langmuir model that was reported in detail in our paper (K. Tawa, D. Yao and W. Knoll, in preparation) was used for the fitting routine. This model describes the kinetics using two components, the main target DNA (m) and a minority species (i) competing for the identical probe sites but with a substantially reduced binding affinity. This then results in an association process that is described by
Figure 2. Hybridization and dissociation measured by SPFS (circles) and fitted curves calculated by the extended Langmuir adsorption model (equations 1 and 2): (A) P2–T1(TG15); (B) P2–T1(TG13); (C) P2–T1(TG09); (D) P2–T1(TG02).
I(t) = I(c0){Am(c0) + koffm}t)] + +koffi}t)]}1
and a dissociation process given by
I(t) = I(c0){Am(c0)exp(-koffmt) + exp(–koffit)}2
where a is the mole fraction of the main part,
Am(c0) = 1/{1 + /},
Am(c0) = Am'(c0)/
and
1 – Am(c0) = Ai(c0) = Ai'(c0)/.
The fitted curves are also given in Figure 2 and the corresponding values are collected in Table 2. The correlation values of the fitted curves for hybridization and dissociation are very good, with R2 > 0.98 for each sample. As concluded from the value of 1 – Am(c0), the minority species contributes only 4–13% to the overall kinetic curves and can be seen, in particular, by its contribution to the dissociation being responsible for the rapid decrease of the fluorescence intensity at early times after the rinsing was started. Hence we discuss only the main components, because their values correspond to the hybridization for the given DNA sequences (K. Tawa, D. Yao and W. Knoll, in preparation). The affinity for duplex formation clearly depends on the position of the mismatching base pair. The following parameters are all the same among the four samples studied: the number of hydrogen bonds, the kind of mismatching base pair (T–G) (25) and nearest-neighbour pairs to the mismatching base pair (T–A and C–G), except for the P2–T1(TG15) for which the mismatching base pair is located at the 3' end of the target. The affinity constants of P2–T1(TG09) and P2–T1(TG02) are KA = 1.5 x 108 M–1 and KA = 1.4 x 108 M–1, respectively, and hence are considerably smaller than the other two (KA = 4.8 x 108 M–1 and KA = 2.8 x 108 M–1, respectively). The hybridization and dissociation rate constants indicate the same tendency as the affinity constants, i.e. kon for P2–T1(TG09) and P2–T1(TG02) is smaller than that of the double strands that incorporate the mismatch near the substrate, i.e. P2–T1(TG15) and P2–T1(TG13), with koff showing the opposite trend. This means that for the P2–T1(TG02) pair with the mismatching base pair located near the 5' end of the target DNA, the mismatching base pair has the strongest destabilizing effect on the hybrid.
Table 2. Fit parameters based on a two-target components model (extended Langmuir model)
It is also interesting to compare the hybridization of MM1 with the behaviour of a fully complementary sequence, i.e. MM0, called P2–T2(15) and P2–T2(14), respectively. T2(15) is a 15mer and T2(14) is a 14mer as described in Table 1, and both are fully complementary to P2. The values for P2–T2(15) evaluated by a Langmuir model are kon = 1.8 x 104 M–1s–1, koff = 1.3 x 10–5 s–1 and KA = 14 x 108 M–1 (Fig. 3A). The P2–T2(14) hybridization parameters are kon = 1.1 x 104 M–1s–1, koff = 2.7 x 10–5 s–1 and KA = 4.0 x 108 M–1 (Fig. 3B). The stability of double strands composed of 15 bases with one mismatch is one order of magnitude smaller than that for P2–T2(15); however, it is of the same order as that of P2–T2(14). The stability of the duplex strongly depends on the number of complementary base pairs, i.e. hydrogen bonds. The results obtained by Okahata et al. (23) also show that the affinity constant of MM0 measured by QCM is one order higher than that of MM1, even though the salt concentration (0.2 M NaCl), temperature (20°C), base-pair unit number (20) and sequence are different from those used in this study. The results of not only the affinity constant but also the hybridization and dissociation rate constants are found to be consistent with our results. Comparing P2–T2(15) with P2–T1(TG15), kon is larger and koff is about twice as small.
Figure 3. Hybridization and dissociation measured by SPFS (circles) and fitted curves calculated by the extended Langmuir adsorption model (equations 1 and 2): (A) P2–T2(15); (B) P2–T2(14).
Dependence of the chemical nature of the mismatching base pair
SantaLucia and coworkers (21,22,25,26) reported the stability for base pairs obtained from statistical simulation results according to the following sequence:
G–C > A–T > G–G > G–T = G–A > T–T = A–A > T–C A–C C–C.
The affinity constant of P2–T1(TG09) shown in Table 2 should be compared with that of P2–T1(TC09) in order to confirm that the kind of mismatching base pair is another important factor in the formation of a duplex. Figure 4A shows the experimental data of the hybridization kinetics of P2 and T1(TC09) and the curve is fitted to the extended Langmuir model (parameters also given in Table 2). The affinity constant of the P2–T1(TC09) pair is KA = 0.53 x 108 M–1 and that of P2–T1(TG09) is KA = 1.5 x 108 M–1. Obviously, the hybridization of P2–T1(TC09) leads to a less stable pair than that of P2–T1(TG09), in accordance with the results obtained by SantaLucia and coworkers (25,26).
Figure 4. Hybridization and dissociation measured by SPFS (circles) and fitted curves calculated by the extended Langmuir adsorption model (equations 1 and 2): (A) P2–T1(TC09); (B) P2–T1(GG14).
However, even if a particular combination of a mismatching base pair resulted in a more stable configuration (according to SantaLucia and coworkers), the difference in the total number of hydrogen bonds induces a different result. As shown in Table 2, P2–T1(GG14) has one mismatching base pair (G–G) and the total number of hydrogen bonds associated with hybridization is one less than that in the P2–T1(TG13) hybrid owing to the replacement of a mismatching pair. Figure 4B shows the association and dissociation process of the P2–T1(GG14) hybrid and Table 2 also shows the fitting result. Larger kon and KA values and a smaller koff value are found for the P2–T1(TG13) hybrid compared with those for P2–T1(GG14). The affinity constant of P2–T1(GG14) is KA = 0.82 x 108 M–1 and that of P2–T1(TG13) is KA = 2.8 x 108 M–1. This indicates that the duplex is becoming more stable as the total number of hydrogen bonds increases, because the mismatching base pair for both is located at almost the same position in the sequence, i.e. near the 3' end of the target. The nearest-neighbour base pairs T–A and T–A may also contribute to the smaller affinity of P2–T2(GG14), because they may interact more weakly than the G–C and A–T pairs in P2–T1(TG13).
The effect of a mismatching base pair at the 3' end of the target
Finally, the hybridization process in the fully complementary sequences, P2–T2(14) hybrid, is compared with the hybridization of the MM1 duplex in P2–T1(TG15) in order to discuss the effect of the end base pair. T2(14) and T1(TG15) have the same sequence except for the 15th base unit counted from the 5' end, and both are matching with P2 base units. The only difference is the 15th base-unit position of the target which is a blank for T2(14), whereas for T1(TG15) there is a T–G forming a mismatching base pair with P2. Also for P2–T2(14), the kinetics curves were fitted by the extended Langmuir model. The fitting values such as the rate constants of hybridization and dissociation for P2–T2(14) are Am = 0.95, konm = 1.1 x 104 M–1 s–1, koffm = 2.7 x 10–5 s–1, KA =4.0 x 108 M–1, Ai =0.05, koni = 1.2 x 104 M–1s–1 and koffi = 100 x 10–5 s–1. The result of the comparison indicates that the duplex of P2–T1(TG15) is slightly more stable than that of P2–T2(14) (Fig. 3B). The affinity constant of P2–T1(TG15) is KA = 4.8 x 108 M–1. As found in the order of the strength of base pairs reported by SantaLucia and coworkers (21,22,25), the T–G mismatching base pair can contribute to the interaction between probe and target oligonucleotides and to the stability of the duplex as a whole despite the fact that it is not of the Watson–Crick character.
CONCLUSION
By using SPFS, we have demonstrated a sensor platform that allows for a very sensitive monitoring of the hybridization reactions between target oligonucleotides from solution labelled with fluorescent molecules and their complementary probe strands at the surface. The kinetic measurements of the association and dissociation processes, respectively, provide the rate constants kon and koff, as well as the affinity constant KA. The chemical nature of the mismatching base pairs contributes to the hybridization kinetics. T–G mismatching base pairs produce a more stable duplex than a T–C pair. This result is in accordance with the result of the simulation of hybridization kinetics in solution. It will be useful information for the development of sensors not only for mutation detection but also for a wider range of molecular recognition studies. Furthermore, a double strand is found to be more destabilized if a mismatching base pair between the target DNA and the probe DNA is located farther away from the solid sensor surface, i.e. facing the PBS solution. This difference in the stability of a duplex due to the mismatch position is also valuable information for studies aiming at detecting mutants and it can be expected to help in the development of optical devices for medical applications. We should point out that these results largely coincide with studies in solutions and/or theoretical predictions but it is the first time that these details of hybridization reactions have been studied experimentally for surface-attached probe strands.
ACKNOWLEDGEMENTS
K.T. acknowledges the Alexander von Humboldt Foundation for a postdoctoral fellowship. Part of this work was supported by an EU grant (QLKI-2000-31658, DNA-Track).
REFERENCES
Knight,J. (2001) Geneticists’ work in disarray as DNA-chip producer pulls the plug. Nature, 414, 135–136.
Halliwell,C.M. and Cass,A.E.G. (2001) A factorial analysis of silanization conditions for the immobilization of oligonucleotides on glass surfaces. Anal. Chem., 73, 2476–2483.
Pirrung,M.C. (2002) How to make a DNA chip. Angew. Chem. Int. Ed. Engl., 41, 1276.
Service,R.F. (1988) DNA analysis: microchip arrays put DNA on the spot. Science, 282, 396–399.
John,D.M. and Weeks,K.M. (2002) Chemical interrogation of mismatches in DNA–DNA and DNA–RNA duplexes under nonstringent conditions by selective 2'-amine acylation. Biochemistry, 41, 6866–6874.
Kirpekar,F., Berkenkamp,S. and Hillenkamp,F. (1999) Detection of double-stranded DNA by IR- and UV-MALDI mass spectrometry. Anal. Chem., 71, 2334–2339.
Sugimoto,N., Satoh,N., Yasuda,K. and Nakano,S. (2001) Sstabilization factors affecting duplex formation of peptide nucleic acid with DNA. Biochemistry, 40, 8444–8451.
Ratilainen,T., Holmén,A., Tuite,E., Nielsen,P.E. and Norden,B. (2000) Thermodynamics of sequence-specific binding of PNA to DNA. Biochemistry, 39, 7781–7791.
Jensen,K.K., ?rum,H., Nielsen,P.E. and Nordén,B. (1997) Kinetics for hybridization of peptide nucleic acids (PNA) with DNA and RNA studied with the BIAcore technique. Biochemistry, 36, 5072–5077.
Wang,J., Nielsen,P.E., Jiang,M., Cai,X.H., Fernandes,J.R., Grant,D.H., Ozsoz,M., Beglieter,A. and Mowat,M. (1997) Mismatch-sensitive hybridization detection by peptide nucleic acids immobilized on a quartz crystal microbalance. Anal. Chem., 69, 5200–5202.
Matsuno,H., Niikura,K. and Okahata,Y. (2001) Design and characterization of asparagine- and lysine-containing alanine-based helical peptides that bind selectively to A·T base pairs of oligonucleotides immobilized on a 27 MHz quartz crystal microbalance. Biochemistry, 40, 3615–3622.
Yamaguchi,S., Shimomura,T., Tatsuma,T. and Oyama,N. (1993) Adsorption, immobilization and hybridization of DNA studied by the use of quartz crystal oscillators. Anal. Chem., 65, 1925–1927.
Jordan,C.E., Frutos,A.G., Thiel,A.J.,and Corn,R.M. (1997) Surface plasmon resonance imaging measurements of DNA hybridization adsorption and streptavidin/DNA multilayer formation at chemically modified gold surfaces. Anal. Chem., 69, 4939–4947.
Watts,H.J., Yeung,D. and Parkes,H. (1995) Real-time detection and quantification of DNA hybridization by an optical biosensor. Anal. Chem., 67, 4283–4289.
Jost,J.P., Munch,O. and Andersson,T. (1991) Study of protein–DNA interactions by surface plasmon resonance (real time kinetics). Nucleic Acids Res., 19, 2788–2788.
Knoll,W. (1998) Interfaces and thin films as seen by bound electromagnetic waves. Annu. Rev. Phys. Chem., 49, 569–638.
Liebermann,T., Knoll,W., Sluka,P. and Herrmann,R. (2000) Complement hybridization from solution to surface-attached probe-oligonucleotides observed by surface-plasmon-field-enhanced fluorescence spectroscopy. Colloids Surf. A, 169, 337–350.
Liebermann,T. and Knoll,W. (2000) Surface-plasmon field-enhanced fluorescence spectroscopy. Colloids Surf. A, 177, 115–130.
Kambhampati,D., Nielsen,P.E. and Knoll,W. (2001) Investigating the kinetics of DNA–DNA and PNA–DNA interactions using surface plasmon resonance-enhanced fluorescence spectroscopy. Biosens. Bioelectron., 16, 1109–1118.
Pierpaoli,E.V., Gisler,S.M. and Christen P. (1998) Sequence-specific rates of interaction of target peptides with the molecular chaperones DnaK and DnaJ. Biochemistry, 37, 16741–16748.
Peyret,N., Seneviratne,P.A., Allawi,H.T. and SantaLucia,J.,Jr (1999) Nearest-neighbor thermodynamics and NMR of DNA sequences with internal A·A, C·C, G·G and T·T mismatches. Biochemistry, 38, 3468–3477.
Allawi,H.T. and SantaLucia,J.,Jr (1998) Thermodynamics of internal C.T mismatches in DNA. Nucleic Acids Res., 26, 2694–2701.
Okahata,Y., Kawase,M., Niikura,K., Ohtake,F., Furusawa,H. and Ebara,Y. (1998) Kinetic measurements of DNA hybridization on an oligonucleotide-immobilized 27-MHz quartz crystal microbalance. Anal. Chem., 70, 1288–1296.
Persson,B., Stenhag,K., Nilsson,P., Larsson,A., Uhlén,M. and Nygren,P.-A. (1997) Analysis of oligonucleotide probe affinities using surface plasmon resonance: a means for mutational scanning. Anal. Biochem., 246, 34–44.
Allawi,H.T.and SantaLucia,J.,Jr (1997) Thermodynamics and NMR of internal G·T mismatches in DNA. Biochemistry, 36, 10581–10594.
Allawi,H.T. and SantaLucia,J.,Jr (1998) Nearest-neighbor thermodynamics of internal A·C mismatches in DNA: sequence dependence and pH effects. Biochemistry, 37, 9435–9444.(Keiko Tawa1,2 and Wolfgang Knoll*,1)
*To whom correspondence should be addressed. Tel: +49 6131 379160; Fax: +49 6131 379360; Email: knoll@mpip-mainz.mpg.de
ABSTRACT
Two single-stranded DNAs consisting of complementary base pairs except for one mismatching base pair (MM1) can form double-stranded DNA by molecular recognition. This type of duplex is not as stable as that formed by MM0. In order to add to a better understanding of the physical mechanism of the hybridization and dissociation processes at sensor (chip) surfaces, we studied the kinetics of the MM1 hybridization by surface plasmon fluorescence spectroscopy. Target DNA strands labelled with a fluorescent molecule Cy5 at the 5' end and hybridizing with the surface-attached probe DNA can be excited by the strong optical field of a surface plasmon resonance mode. The emitted fluorescence can be detected with high sensitivity. The affinity of a duplex was found to depend on the chemical nature, i.e. G–G, G–T etc., and on the position of the mismatching base pair along the 15mer duplex.
INTRODUCTION
The need for optical techniques sensitive enough to detect mismatching base pairs was greatly increased by the urgent request to identify point mutations associated with cancer and other diseases. Especially, technologies based on the DAN-chip platform (1–4) using hybridization strategies between a single-stranded probe DNA immobilized on the sensor surface and a complementary target strand binding from solution are required. Therefore not only DNA–DNA but also peptide nucleic acid (PNA)–DNA hybridization reactions have been studied by various groups as a very important research subject, applying a broad range of experimental techniques (5–9). The usual methods employed to observe such hybridization reactions are based on monitoring the mass deposition measured with the quartz crystal microbalance (QCM) (10–12) or recording the change in the optical thickness at the surface detected by surface plasmon spectroscopy (SPS) (13–16). Recently we introduced surface plasmon fluorescence spectroscopy (SPFS) as a technique that offers higher detection limits and higher sensitivity than the above mentioned methods (16,17). In order to evaluate the kinetics of DNA–DNA hybridization without intermolecular coulombic interactions between neighbouring binding sites, our strategy is to limit the probe DNA to a maximum density of 2.5 x 1012 cm–2, which was shown to be still near the limit of infinite dilution by binding the probes via their biotin end to a streptavidin monolayer. In such a dilute surface, hybridization for 15 bp between probe and target is difficult to observe by SPR quantitatively; however, a quantitative analysis is possible by the in situ recording of the kinetics of the association (hybridization) and dissociation process (18) by SPFS using target oligo-DNA chemically labelled with a fluorescent molecule. The kinetic parameters obtained by SPFS can be theoretically analyzed by a Langmuir adsorption model, i.e. the rate constants for hybridization and dissociation can be evaluated for any hybridization systems. This is a real advantage compared with other experimental techniques; hence we consider SPFS as one of the most powerful experimental methods even though the target DNA has to be modified with a fluorescent molecule. However, this label does not modify the measured kinetic parameters as we could show independently by recent measurements in which the chromophore was attached to the probe strand and unlabelled target oligonucleotides hybridized with virtually identical rate constants. Moreover, a recent novel label-free detection principle based on surface plasmon diffraction which has a higher sensitivity than classical SPR, and hence allows for the recording of oligonucleotide hybridization and dissociation (though with a weaker signal-to-noise ratio compared with SPFS) without a chromophore attached to either single strand, confirmed the negligible influence of the presence of the fluorophore in the measured kinetic rate constants and affinity constants.
So far, we have shown that mismatch discrimination is easily achieved with high sensitivity (19). However, details of the effect of mismatching (20–22) and matching (K. Tawa, D. Yao and W. Knoll, in preparation) base pairs were not investigated. Therefore more experimental studies concerned with the type and position of a mismatching base pair in oligo-DNA sequences were considered to be necessary in order to clarify the mechanism of hybridization. Here, we focus on how the position and the chemical nature (combination) of a mismatching base pair can contribute to the kinetics of hybridization between 15mer oligonucleotide strands.
MATERIALS AND METHODS
Oligo-DNA samples (purification by HPSF) were obtained from MWG-biotech AG. The sequence of the various base units is collected in Table 1. The target oligonucleotides composed of bases complementary to the probe DNA are called MM0. The target DNA with one mismatching base is called MM1. Each oligo-DNA in the MM1 series has 15 base units with different positions of the mismatching base pair and different chemical nature (combinations) of base mismatches. The details of SPFS were reported in previous papers (16–19, K. Tawa, D. Yao and W. Knoll, in preparation). Briefly, the set-up that we use is based on essential modules of a ‘normal’ surface plasmon spectrometer as shown in Figure 1A (17,18). A He–Ne laser operating at 632.8 nm passes an optical chopper (used also as the reference for the lock-in amplifier) and two polarizers for intensity and polarization control. By using a –2 goniometer, the light reflected at the substrate covered with a thin gold layer is collected through a lens and monitored by a photodiode. If the reflectivity is observed at a fixed angle of incidence, any change of the interfacial architecture (e.g. induced by adsorption or desorption processes) can be quantified by evaluating the resulting change of the reflected intensity (16). The intensity of surface plasmon excited fluorescence was also measured by a photomultiplier at a fixed angle of excitation for which the reflectivity was R = 0.30. The time-dependent fluorescence intensity observed by SPFS corresponds to the kinetics of hybridization and dissociation between probe and target DNAs (17).
Table 1. Sequences of the oligo-DNA samples employed
Figure 1. (A) SPFS set-up. (B) Sample layers used for the DNA–DNA hybridization studies.
As has been discussed in a previous paper (18), the optical field which is strongly enhanced at the surface plasmon (PSP) resonance angle can be used to excite fluorescence molecules attached to the surface-bound analyte. The angular characteristic of the fluorescence intensity then follows the angular dependence of the optical field at the interface. While in the case of total internal reflection only an intensity enhancement of a factor of 4 can be achieved at the critical angle of total internal reflection, enhancement factors of up to about 20 can be achieved for gold. Fluorescent molecules that are within the evanescent field of the PSP, which decays normal to the interface with a penetration depth of Lz = 150 nm, can be excited by this surface plasmon mode. The emission is monitored by a photomultiplier (after passing through a set of interference filters) mounted to the goniometer part that rotates with , as does the coupling prism (see Fig. 1).
In order to achieve an optimized fluorescence signal and not to lose too much intensity by quenching to the thin gold layer (F?rster transfer), the chromophores have to be sufficiently separated from the gold surface. The relevant length scale is given by the F?rster radius which amounts to 5–7 nm. The molecular architecture of the interfacial layer, including the oligo-DNA layer prepared on the substrate, was as in our previous studies (as shown in Fig. 1B) (16–19). We achieve this by building the biotin/streptavidin architecture that ensures the required separation of the chromophores attached to the target DNA strands. First, a mixed self-assembled monolayer (SAM) of a biotinylated thiol derivative and a short-chain OH-terminated thiol used as a diluent at a mixing ratio of 1:9 was prepared on the gold substrates in order to optimize the formation of a streptavidin monolayer via the specific recognition to the biotin moieties. Then, the next layer of oligonucleotide catcher probes with a biotin group at their 5' end are organized in a rather ordered way at that interface. The density of probe DNA thus obtained corresponds to 2.5 x 1012 cm–2. The probe and target sequences used in this study are summarized in Table 1. All target sequences are labelled by a fluorophore called Cy5 at their 5' end. The sample cell employed is made from quartz glass and contains in- and outlets for the various solutions. It is sealed via a Viton O-ring against the gold-coated high refractive glass slide which is index-matched to the prism.
All solutions used in this study were prepared using phosphate-buffered saline (PBS) tablets (Aldrich-Sigma) with an Na+ concentration of 0.137 M. All measurements were performed at room temperature (22–23°C). The concentration of the target DNA solution was 0.25 μM for the measurement of the hybridization kinetics, and the solutions were applied via injection of 200–300 μl with a syringe. Once a stable surface coverage of bound target was reached, the dissociation process was initiated by rinsing pure PBS solution with a peristaltic pump through the cell.
RESULTS AND DISCUSSION
Hybridization kinetics: mismatching base-pair position dependence
Major factors that were reported to contribute to the interfacial binding process are the ionic strength and the pH in the buffer solution (23), the temperature (23,24), the sequence length and the chemical nature (combination) of base units (24). The comparison of the kinetics of the P2–T1(TG15), P2–T1(TG13), P2–T1(TG09) and P2–T1(TG02) probe–target pairs should be helpful for the understanding of the mismatching base-pair position dependence of the hybridization process. The corresponding kinetic curves are shown in Figure 2. As it turned out, a simple Langmuir model describing the association as well as the dissociation step by a single rate constant kon and koff, respectively, could not fit the data appropriately. Therefore an extended Langmuir model that was reported in detail in our paper (K. Tawa, D. Yao and W. Knoll, in preparation) was used for the fitting routine. This model describes the kinetics using two components, the main target DNA (m) and a minority species (i) competing for the identical probe sites but with a substantially reduced binding affinity. This then results in an association process that is described by
Figure 2. Hybridization and dissociation measured by SPFS (circles) and fitted curves calculated by the extended Langmuir adsorption model (equations 1 and 2): (A) P2–T1(TG15); (B) P2–T1(TG13); (C) P2–T1(TG09); (D) P2–T1(TG02).
I(t) = I(c0){Am(c0) + koffm}t)] + +koffi}t)]}1
and a dissociation process given by
I(t) = I(c0){Am(c0)exp(-koffmt) + exp(–koffit)}2
where a is the mole fraction of the main part,
Am(c0) = 1/{1 + /},
Am(c0) = Am'(c0)/
and
1 – Am(c0) = Ai(c0) = Ai'(c0)/.
The fitted curves are also given in Figure 2 and the corresponding values are collected in Table 2. The correlation values of the fitted curves for hybridization and dissociation are very good, with R2 > 0.98 for each sample. As concluded from the value of 1 – Am(c0), the minority species contributes only 4–13% to the overall kinetic curves and can be seen, in particular, by its contribution to the dissociation being responsible for the rapid decrease of the fluorescence intensity at early times after the rinsing was started. Hence we discuss only the main components, because their values correspond to the hybridization for the given DNA sequences (K. Tawa, D. Yao and W. Knoll, in preparation). The affinity for duplex formation clearly depends on the position of the mismatching base pair. The following parameters are all the same among the four samples studied: the number of hydrogen bonds, the kind of mismatching base pair (T–G) (25) and nearest-neighbour pairs to the mismatching base pair (T–A and C–G), except for the P2–T1(TG15) for which the mismatching base pair is located at the 3' end of the target. The affinity constants of P2–T1(TG09) and P2–T1(TG02) are KA = 1.5 x 108 M–1 and KA = 1.4 x 108 M–1, respectively, and hence are considerably smaller than the other two (KA = 4.8 x 108 M–1 and KA = 2.8 x 108 M–1, respectively). The hybridization and dissociation rate constants indicate the same tendency as the affinity constants, i.e. kon for P2–T1(TG09) and P2–T1(TG02) is smaller than that of the double strands that incorporate the mismatch near the substrate, i.e. P2–T1(TG15) and P2–T1(TG13), with koff showing the opposite trend. This means that for the P2–T1(TG02) pair with the mismatching base pair located near the 5' end of the target DNA, the mismatching base pair has the strongest destabilizing effect on the hybrid.
Table 2. Fit parameters based on a two-target components model (extended Langmuir model)
It is also interesting to compare the hybridization of MM1 with the behaviour of a fully complementary sequence, i.e. MM0, called P2–T2(15) and P2–T2(14), respectively. T2(15) is a 15mer and T2(14) is a 14mer as described in Table 1, and both are fully complementary to P2. The values for P2–T2(15) evaluated by a Langmuir model are kon = 1.8 x 104 M–1s–1, koff = 1.3 x 10–5 s–1 and KA = 14 x 108 M–1 (Fig. 3A). The P2–T2(14) hybridization parameters are kon = 1.1 x 104 M–1s–1, koff = 2.7 x 10–5 s–1 and KA = 4.0 x 108 M–1 (Fig. 3B). The stability of double strands composed of 15 bases with one mismatch is one order of magnitude smaller than that for P2–T2(15); however, it is of the same order as that of P2–T2(14). The stability of the duplex strongly depends on the number of complementary base pairs, i.e. hydrogen bonds. The results obtained by Okahata et al. (23) also show that the affinity constant of MM0 measured by QCM is one order higher than that of MM1, even though the salt concentration (0.2 M NaCl), temperature (20°C), base-pair unit number (20) and sequence are different from those used in this study. The results of not only the affinity constant but also the hybridization and dissociation rate constants are found to be consistent with our results. Comparing P2–T2(15) with P2–T1(TG15), kon is larger and koff is about twice as small.
Figure 3. Hybridization and dissociation measured by SPFS (circles) and fitted curves calculated by the extended Langmuir adsorption model (equations 1 and 2): (A) P2–T2(15); (B) P2–T2(14).
Dependence of the chemical nature of the mismatching base pair
SantaLucia and coworkers (21,22,25,26) reported the stability for base pairs obtained from statistical simulation results according to the following sequence:
G–C > A–T > G–G > G–T = G–A > T–T = A–A > T–C A–C C–C.
The affinity constant of P2–T1(TG09) shown in Table 2 should be compared with that of P2–T1(TC09) in order to confirm that the kind of mismatching base pair is another important factor in the formation of a duplex. Figure 4A shows the experimental data of the hybridization kinetics of P2 and T1(TC09) and the curve is fitted to the extended Langmuir model (parameters also given in Table 2). The affinity constant of the P2–T1(TC09) pair is KA = 0.53 x 108 M–1 and that of P2–T1(TG09) is KA = 1.5 x 108 M–1. Obviously, the hybridization of P2–T1(TC09) leads to a less stable pair than that of P2–T1(TG09), in accordance with the results obtained by SantaLucia and coworkers (25,26).
Figure 4. Hybridization and dissociation measured by SPFS (circles) and fitted curves calculated by the extended Langmuir adsorption model (equations 1 and 2): (A) P2–T1(TC09); (B) P2–T1(GG14).
However, even if a particular combination of a mismatching base pair resulted in a more stable configuration (according to SantaLucia and coworkers), the difference in the total number of hydrogen bonds induces a different result. As shown in Table 2, P2–T1(GG14) has one mismatching base pair (G–G) and the total number of hydrogen bonds associated with hybridization is one less than that in the P2–T1(TG13) hybrid owing to the replacement of a mismatching pair. Figure 4B shows the association and dissociation process of the P2–T1(GG14) hybrid and Table 2 also shows the fitting result. Larger kon and KA values and a smaller koff value are found for the P2–T1(TG13) hybrid compared with those for P2–T1(GG14). The affinity constant of P2–T1(GG14) is KA = 0.82 x 108 M–1 and that of P2–T1(TG13) is KA = 2.8 x 108 M–1. This indicates that the duplex is becoming more stable as the total number of hydrogen bonds increases, because the mismatching base pair for both is located at almost the same position in the sequence, i.e. near the 3' end of the target. The nearest-neighbour base pairs T–A and T–A may also contribute to the smaller affinity of P2–T2(GG14), because they may interact more weakly than the G–C and A–T pairs in P2–T1(TG13).
The effect of a mismatching base pair at the 3' end of the target
Finally, the hybridization process in the fully complementary sequences, P2–T2(14) hybrid, is compared with the hybridization of the MM1 duplex in P2–T1(TG15) in order to discuss the effect of the end base pair. T2(14) and T1(TG15) have the same sequence except for the 15th base unit counted from the 5' end, and both are matching with P2 base units. The only difference is the 15th base-unit position of the target which is a blank for T2(14), whereas for T1(TG15) there is a T–G forming a mismatching base pair with P2. Also for P2–T2(14), the kinetics curves were fitted by the extended Langmuir model. The fitting values such as the rate constants of hybridization and dissociation for P2–T2(14) are Am = 0.95, konm = 1.1 x 104 M–1 s–1, koffm = 2.7 x 10–5 s–1, KA =4.0 x 108 M–1, Ai =0.05, koni = 1.2 x 104 M–1s–1 and koffi = 100 x 10–5 s–1. The result of the comparison indicates that the duplex of P2–T1(TG15) is slightly more stable than that of P2–T2(14) (Fig. 3B). The affinity constant of P2–T1(TG15) is KA = 4.8 x 108 M–1. As found in the order of the strength of base pairs reported by SantaLucia and coworkers (21,22,25), the T–G mismatching base pair can contribute to the interaction between probe and target oligonucleotides and to the stability of the duplex as a whole despite the fact that it is not of the Watson–Crick character.
CONCLUSION
By using SPFS, we have demonstrated a sensor platform that allows for a very sensitive monitoring of the hybridization reactions between target oligonucleotides from solution labelled with fluorescent molecules and their complementary probe strands at the surface. The kinetic measurements of the association and dissociation processes, respectively, provide the rate constants kon and koff, as well as the affinity constant KA. The chemical nature of the mismatching base pairs contributes to the hybridization kinetics. T–G mismatching base pairs produce a more stable duplex than a T–C pair. This result is in accordance with the result of the simulation of hybridization kinetics in solution. It will be useful information for the development of sensors not only for mutation detection but also for a wider range of molecular recognition studies. Furthermore, a double strand is found to be more destabilized if a mismatching base pair between the target DNA and the probe DNA is located farther away from the solid sensor surface, i.e. facing the PBS solution. This difference in the stability of a duplex due to the mismatch position is also valuable information for studies aiming at detecting mutants and it can be expected to help in the development of optical devices for medical applications. We should point out that these results largely coincide with studies in solutions and/or theoretical predictions but it is the first time that these details of hybridization reactions have been studied experimentally for surface-attached probe strands.
ACKNOWLEDGEMENTS
K.T. acknowledges the Alexander von Humboldt Foundation for a postdoctoral fellowship. Part of this work was supported by an EU grant (QLKI-2000-31658, DNA-Track).
REFERENCES
Knight,J. (2001) Geneticists’ work in disarray as DNA-chip producer pulls the plug. Nature, 414, 135–136.
Halliwell,C.M. and Cass,A.E.G. (2001) A factorial analysis of silanization conditions for the immobilization of oligonucleotides on glass surfaces. Anal. Chem., 73, 2476–2483.
Pirrung,M.C. (2002) How to make a DNA chip. Angew. Chem. Int. Ed. Engl., 41, 1276.
Service,R.F. (1988) DNA analysis: microchip arrays put DNA on the spot. Science, 282, 396–399.
John,D.M. and Weeks,K.M. (2002) Chemical interrogation of mismatches in DNA–DNA and DNA–RNA duplexes under nonstringent conditions by selective 2'-amine acylation. Biochemistry, 41, 6866–6874.
Kirpekar,F., Berkenkamp,S. and Hillenkamp,F. (1999) Detection of double-stranded DNA by IR- and UV-MALDI mass spectrometry. Anal. Chem., 71, 2334–2339.
Sugimoto,N., Satoh,N., Yasuda,K. and Nakano,S. (2001) Sstabilization factors affecting duplex formation of peptide nucleic acid with DNA. Biochemistry, 40, 8444–8451.
Ratilainen,T., Holmén,A., Tuite,E., Nielsen,P.E. and Norden,B. (2000) Thermodynamics of sequence-specific binding of PNA to DNA. Biochemistry, 39, 7781–7791.
Jensen,K.K., ?rum,H., Nielsen,P.E. and Nordén,B. (1997) Kinetics for hybridization of peptide nucleic acids (PNA) with DNA and RNA studied with the BIAcore technique. Biochemistry, 36, 5072–5077.
Wang,J., Nielsen,P.E., Jiang,M., Cai,X.H., Fernandes,J.R., Grant,D.H., Ozsoz,M., Beglieter,A. and Mowat,M. (1997) Mismatch-sensitive hybridization detection by peptide nucleic acids immobilized on a quartz crystal microbalance. Anal. Chem., 69, 5200–5202.
Matsuno,H., Niikura,K. and Okahata,Y. (2001) Design and characterization of asparagine- and lysine-containing alanine-based helical peptides that bind selectively to A·T base pairs of oligonucleotides immobilized on a 27 MHz quartz crystal microbalance. Biochemistry, 40, 3615–3622.
Yamaguchi,S., Shimomura,T., Tatsuma,T. and Oyama,N. (1993) Adsorption, immobilization and hybridization of DNA studied by the use of quartz crystal oscillators. Anal. Chem., 65, 1925–1927.
Jordan,C.E., Frutos,A.G., Thiel,A.J.,and Corn,R.M. (1997) Surface plasmon resonance imaging measurements of DNA hybridization adsorption and streptavidin/DNA multilayer formation at chemically modified gold surfaces. Anal. Chem., 69, 4939–4947.
Watts,H.J., Yeung,D. and Parkes,H. (1995) Real-time detection and quantification of DNA hybridization by an optical biosensor. Anal. Chem., 67, 4283–4289.
Jost,J.P., Munch,O. and Andersson,T. (1991) Study of protein–DNA interactions by surface plasmon resonance (real time kinetics). Nucleic Acids Res., 19, 2788–2788.
Knoll,W. (1998) Interfaces and thin films as seen by bound electromagnetic waves. Annu. Rev. Phys. Chem., 49, 569–638.
Liebermann,T., Knoll,W., Sluka,P. and Herrmann,R. (2000) Complement hybridization from solution to surface-attached probe-oligonucleotides observed by surface-plasmon-field-enhanced fluorescence spectroscopy. Colloids Surf. A, 169, 337–350.
Liebermann,T. and Knoll,W. (2000) Surface-plasmon field-enhanced fluorescence spectroscopy. Colloids Surf. A, 177, 115–130.
Kambhampati,D., Nielsen,P.E. and Knoll,W. (2001) Investigating the kinetics of DNA–DNA and PNA–DNA interactions using surface plasmon resonance-enhanced fluorescence spectroscopy. Biosens. Bioelectron., 16, 1109–1118.
Pierpaoli,E.V., Gisler,S.M. and Christen P. (1998) Sequence-specific rates of interaction of target peptides with the molecular chaperones DnaK and DnaJ. Biochemistry, 37, 16741–16748.
Peyret,N., Seneviratne,P.A., Allawi,H.T. and SantaLucia,J.,Jr (1999) Nearest-neighbor thermodynamics and NMR of DNA sequences with internal A·A, C·C, G·G and T·T mismatches. Biochemistry, 38, 3468–3477.
Allawi,H.T. and SantaLucia,J.,Jr (1998) Thermodynamics of internal C.T mismatches in DNA. Nucleic Acids Res., 26, 2694–2701.
Okahata,Y., Kawase,M., Niikura,K., Ohtake,F., Furusawa,H. and Ebara,Y. (1998) Kinetic measurements of DNA hybridization on an oligonucleotide-immobilized 27-MHz quartz crystal microbalance. Anal. Chem., 70, 1288–1296.
Persson,B., Stenhag,K., Nilsson,P., Larsson,A., Uhlén,M. and Nygren,P.-A. (1997) Analysis of oligonucleotide probe affinities using surface plasmon resonance: a means for mutational scanning. Anal. Biochem., 246, 34–44.
Allawi,H.T.and SantaLucia,J.,Jr (1997) Thermodynamics and NMR of internal G·T mismatches in DNA. Biochemistry, 36, 10581–10594.
Allawi,H.T. and SantaLucia,J.,Jr (1998) Nearest-neighbor thermodynamics of internal A·C mismatches in DNA: sequence dependence and pH effects. Biochemistry, 37, 9435–9444.(Keiko Tawa1,2 and Wolfgang Knoll*,1)