2944
J. Phys. Chem. B 2009, 113, 2944–2951
Effects for the Incorporation of Five-atom Thioacetamido Nucleic Acid (TANA) Backbone on Hybridization Thermodynamics and Kinetics of DNA Duplexes Harleen Kaur,† Amit Arora,† K. Gogoi,‡ P. Solanke,‡ Anita D. Gunjal,‡ Vaijayanti A. Kumar,*,‡ and Souvik Maiti*,† Institute of Genomics and IntegratiVe Biology, CSIR, Mall Road, Delhi 110 007, India, and DiVision of Organic Chemistry, National Chemical Laboratory, Dr. Homi Bhabha Road, Pune 411008, India ReceiVed: October 3, 2008; ReVised Manuscript ReceiVed: December 1, 2008
Thermodynamic studies of nucleic acids serve not only to widen our understanding on the nature and strength of forces that stabilize nucleic acids in a myriad of structural states they assume but also to facilitate the development of databases that could be used to predict the stability and selectivity of probe/primer-sets required in a broad range of nucleic acid-based diagnostic and therapeutic protocols. In the current study, we investigated the effect of a novel, backbone-modified “thioacetamido linkage” (TANA) on thermodynamics of hybridization, binding kinetics, and conformation of a DNA duplex. The modification comprises of an extended five-atom amide (N3′-CO-CH2-S-CH2) linker, as opposed to four-atom phosphodiester linker backbone present in DNA. One to three TANA-substitutions have been introduced in the linker backbone of two thymidine residues of one of the strand of the DNA duplex. Using spectroscopic and calorimetric techniques, we observed that TANA destabilizes the DNA helix by lowering the favorable enthalpy parameter of duplex formation. TANA · DNA duplexes were found to adopt a conformation toward an A-type duplex as shown by circular dichroism spectroscopy studies. Analysis of differential scanning calorimetry data indicated a nonzero heat capacity change, ∆Cp, accompanying the duplex formation. The average ∆Cp change per duplex was found to be 832.5 cal mol-1 K-1, giving an average base-pair change of 59.5 cal (mol of base pairs)-1 K-1. Hybridization kinetic measurements using surface plasmon resonance indicated a decrease in binding affinity parameter (KA) that originates from higher dissociation rate constants (kd). Furthermore, optical melting studies showed that increasing the number of modifications results in a modest change in the number of counterions taken up during duplex formation. Introduction In the past few decades, huge effort has been invested to unveil the complexities of gene regulatory networks. The discovery of the mechanism wherein regulation occurs through complementary molecules and is encoded within the DNA itself has revolutionalized the arena of modern research and medicine. Efforts are now being channeled to harness the immense potential of short, synthetic oligonucleotides to modulate gene expression by complementary binding to the biologically relevant nucleic acid sequences. Accordingly, a large number of modified nucleotide analogs with unusual chemistries have been developed with properties superior to the native DNA and RNA units. Chemical modifications have been introduced at the heterocyclic bases, the deoxyribose sugar, or to the phosphodiester linkage of DNA strand.1 Backbone substitutions like phosphorothioates,2 peptide nucleic acids,3 ribose sugar substitutions such as 2′-O-alkyl,4 the locked nucleic acid,5 the 2′-Farabinoside,6 and the base alkyl substitutions,7 etc. are among some of the few interesting modifications. Since, the ensemble of atoms constituting DNA or RNA provides a large scope for alteration or attachment of new functional moieties, testing for the functional performance of all these modified oligonucleotides, especially under various experimental settings, is a * To whom correspondence should be addressed. Phone: +91-11-27666156; fax: +91-11-2766-7471; e-mail:
[email protected] (S. M.),
[email protected] (V. A. K.). † Institute of Genomics and Integrative Biology. ‡ National Chemical Laboratory.
tedious task. The knowledge of thermodynamics in this context comes handy and serves to be of fundamental importance. Characterizing the hybridization thermodynamics of modified oligonucleotide provides the fundamental information about the effect of modification on the physicochemical attributes of oligonucleotide such as stability and specificity, which directly correlate with the functional performance of a modified oligonucleotide. The evaluation of thermodynamic parameters allows for accurate prediction about the performance of these modified versions in different biological settings or under different temperature, pH, or ionic contexts. Furthermore, an accurate correlation of thermodynamic parameters with the functional efficiency allows creation of the databases for convenient designing of highly efficient probe/primer-sets required in a broad range of nucleic acid-based diagnostic and therapeutic protocols. We have undertaken this study to characterize the hybridization thermodynamics of a recently developed backbone-modified nucleic acid analog, namely, the thioacetamido nucleic acid (TANA).8 The modification here consists of a backbone with a five-atom amide (N3′-CO-CH2-S-CH2) linker that connects either two thymidine blocks or a thymidine with a cytidine block (Scheme 1). The modification in this case is unusual as the repetitive backbone unit in DNA and almost all commercially available analogs, including PNAs, contain the repetitive unit of six atoms that includes the P, O5′, C5′, C4′, C3′, and O3′ and, thus, a four atom spacer linking the adjacent five-membered rings (O3′, P, O5′, and C5′). Interestingly, analogs with
10.1021/jp808747g CCC: $40.75 2009 American Chemical Society Published on Web 02/06/2009
Energetics of TANA-modified Duplexes SCHEME 1: The molecular structure of five-atom linker thioacetamido nucleic acid (TANA)
backbones one atom shorter than the natural nucleic acids, such as threofuranosyl nucleic acid (TNA),9 lyxopyranosyl nucleic acid (LPNA),10 and glycol nucleic acid (GNA),11 have also been developed and have been found to display effective pairing with DNA and RNA. These observations demonstrate that the duplex formation via Watson-Crick base pairing is not an exclusive attribute of species with backbones composed of six-atom repeat units and that base pairing does not require matching lengths of backbones. In the case of four-atom linker backbone, the effect of amide (-CONR-) linkage on the thermostability of RNA has been extensively characterized and has been shown to be dependent on the type of substituent attached to the nitrogen. It has been shown that maximum thermal stability for the helix could be achieved when R ) H.12 Further, studies revealed that the higher stability of such amide-based linkages arise because of more favorable base-stacking and electrostatic interactions.13 In contrast to the analogs with shortened backbones, only limited information is available about the effect of backbone extensions (i.e., five-atom spacers linking cyclic portions of sugars from adjacent residues) on duplex formation. For the extended backbone systems also, the effect has been characterized exclusively for the linkages carrying amide groups. The few reports based on characterization of these analogs have demonstrated that in comparison to four-atom linkers, the presence of five-atom amide linker in the DNA backbone results in reduced RNA binding,14 whereas another report revealed that the presence of five-atom amide linkers in a DNA leads to slight destabilization of RNA-DNA hybrids.15,16 Yet, another report demonstrated that introduction of five-atom amide linkers in DNA leads to slight destabilization (decreased ∆H) of DNA · RNA helix unless combined with appropriate sugar modifications. Further, this efficiency of hybridization depends on the absolute stereochemistry of the linker moiety.17 The current work highlights the effect of incorporation of five-atom linker TANA on the hybridization thermodynamics and kinetics of DNA · DNA duplex formation. Complete thermodynamic characterization of TANA-modified DNA duplexes, carrying multiple modifications within the thymidine dimer, has been done using spectroscopic and calorimetric analysis. Introduction of this modification in one of the strands of the duplex was found to be accompanied by low thermal (Tm) and low thermodynamic stability (∆G) compared to an unmodified duplex. A detailed analysis showed that increasing the number of TANA-modifications decreases the associated change in the favorable enthalpy, which overrides the effect of decrease in
J. Phys. Chem. B, Vol. 113, No. 9, 2009 2945 TABLE 1: Positioning of TANA Substitutions within the Thymidine Blocks of the Oligonucleotide Are Shown as “tst” nomenclature
sequence (14-mer)
TANA-0 (unmodified oligonucleotide) TANA-1 (single modification) TANA-2 (double modification) TANA-3 (triple modification)
5′- GCTTCCTTCCTTCG -3′ 3′-CGAAGGAAGGAAGC-5′ 5′- GCtstCCTTCCTTCG -3′ 3′-CGAAGGAAGGAAGC-5′ 5′- GCtstCCtstCCTTCG -3′ 3′-CGAAGGAAGGAAGC-5′ 5′- GCtstCCtstCCtstCG -3′ 3′-CGAAGGAAGGAAGC-5′
the unfavorable entropy term, thereby making the overall process of duplex formation energetically unfavorable. Differential scanning calorimetry (DSC) analysis has been performed to evaluate the associated heat capacity change (∆Cp) associated with the individual duplexes and has been used to furnish parameters for a complete thermodynamic profile at 25 °C. In addition to the thermodynamic characterization, the binding kinetic studies have also been undertaken to investigate the underlying causes for TANA-associated DNA duplex destabilization. Hybridization kinetics study for TANA-modified strands to unmodified DNA strand, using surface plasmon resonance (SPR), demonstrates a decrease in the binding affinity by almost an order of 101 with an increase in the number of modifications from zero to three in one of the strands of the duplex. Materials and Methods The hybridization thermodynamics and kinetics of TANAbased oligonucleotides have been studied for a set of four oligonucleotides (14-mer) sequences containing TANA substitutions ranging from zero to three (Table 1). For all four sequences the linker modification (TANA) was introduced between the two thymidylate blocks. The TANA-containing oligonucleotides were synthesized and purified as described by us previously.8 The other unmodified oligonucleotides (HPLC-purified) were purchased from SBS Genetech. The solution concentrations of each of the unmodified oligonucleotide were determined optically at 260 nm and 25 °C using the following molar extinction coefficients: 120 mM-1 cm-1 for d(GCTTCCTTCCTTCG), and 113 mM-1 cm-1 for d(CGAAGGAAGGAAGC). These values were calculated by extrapolation of the tabulated values of the dimer and monomer nucleotides at 25 °C to high temperatures using protocols reported previously.19 For the modified oligonucleotides, the molar absorptivities were assumed to be identical to the DNA oligonucleotides. All the measurements, except kinetics study through SPR, were performed in 10 mM sodium cacodylate buffer (pH 7.4) containing 10 mM sodium chloride. The duplex solutions were prepared by mixing the given template strand bearing varying amount of TANA modifications, with the corresponding unmodified complementary strand in ratio 1:1. Prior to the experiments, the samples were heated to 95 °C for 20 min and then slowly annealed to the starting temperature of the experimental conditions. Circular Dichroism (CD) Spectroscopy. Evaluation of the conformation of each duplex (Table 1) was done by inspection of their CD spectrum at 10 °C. All spectra were recorded using a JASCO J-715 circular dichroism spectropolarimeter equipped with a Julabo heating/cooling system, using a 10 mm quartz cuvette. Following temperature equilibration, ellipticity data was collected between 200 and 350 nm, using a wavelength step of 1 nm. The reported spectra were obtained for a duplex concentration of 5 µM and correspond to the average of three scans.
2946 J. Phys. Chem. B, Vol. 113, No. 9, 2009
Kaur et al.
Temperature-Dependent UV Spectroscopy (UV Melts). Thermal denaturation scans for each of the 14 bp duplexes were obtained with a thermoelectrically controlled Cary 100 (Varian) spectrophotometer, as a function of salt (10-110 mM Na+) concentration. All melting curves for duplex denaturation were collected at a 260 nm wavelength as a function of temperature in the temperature range from 5-75 °C at a heating rate of 0.5 °C/min. To obtain accurate baselines at lower temperature range, the instrument was flushed continuously with nitrogen to avoid any erroneous data collection due to condensation. The mole fraction of the unfolded duplex (R) at different temperatures was evaluated from the absorbance, and ∆A/∆Amax, where ∆A is the change in absorbance at 260 nm at any temperature and ∆Amax is the maximum change recorded at the highest temperature, was also calculated. The plot of mole fraction (R) versus temperature allowed measurement of the melting temperature, Tm, marked by the midpoint of order-disorder transitions. The van’t Hoff enthalpy change (∆HvH) for the duplex formation was obtained from the slope of the 1/Tm versus ln(CT/ 4) plot, according to the following equation for nonselfcomplementary oligonucleotides,20 where
1 R ∆S ) ln(CT /4) + Tm ∆HVH ∆HVH
(1)
Here, R is the universal gas constant (1.986 cal mol-1 K-1), and CT represents the total strand concentration. The slope and the intercept of the plot gave the model-dependent enthalpy and entropy values, respectively. The Tm values for respective curves were identified using dR/dT versus temperature plot. Van’t Hoff enthalpies for duplex formation were also calculated by fitting the shape of each curve to the following equation using the Mathemetica 5.1. The fitting of the curves is done based on the following set of equations: Briefly, the equilibrium constant describing the hybridization of nonself-complementary oligonucleotides is given by:
Keq )
2R (1 - R)2CT
(2)
where R is the fraction of the total DNA strands found in doublestranded form. Solving the equation for R, and choosing the physically realizable root, we have
R)
1 + CTKeq - √1 + 2CTKeq CTKeq
(3)
The van’t Hoff expression for Keq is
Keq ) exp(-∆G°/RT) ) exp(∆H°/RT + ∆S°/R) (4) where the enthalpy (∆H°) and entropy (∆S°) changes refer to hybridization, so both are negative. The melting temperature, Tm was obtained from the dR/dT versus temperature plots of each curve. The absorbance, A260, at temperature T is given by
A260 ) R[A260(Tmin) + mds(T - Tmin)] + (1 - R)[A260(Tmax) - mss(Tmax - T)] (5) where A260(Tmin) and A260(Tmax) are the measured absorbances at temperatures Tmin and Tmax, and mss and mds contain the temperature dependence of the extinction coefficients of the ssDNA and dsDNA, respectively. A260(Tmin) and A260(Tmax) are taken as the average absorbances represented by the intercept for the temperatures at upper and lower baselines, respectively; and mds and mss are the average slopes for the same baselines, respectively. Furthermore, the thermodynamic uptake of counterions, ∆nNa+, associated with the process of duplex formation was obtained according to the equation
∆nNa + )
1.11(∆H/RT2m)∆Tm ∆(ln[Na+])
(6)
where 1.11 is a proportionality constant for converting ionic activity into concentrations and δTm/δ(ln[Na+]) represents the slope of a plot of Tm versus the logarithm of sodium ions (ln[Na+]) at different concentrations (10-110 mM Na+).21 Differential Scanning Calorimetry (DSC Melts). Differential scanning calorimetry was used to monitor the heat capacities of each duplex (50 µM) as a function of temperature (melting curves) using VP-DSC differential scanning calorimeter (Microcal, Inc., Northampton, MA). A typical experiment involved heating of a sample cell containing 0.52 mL of duplex solution and a reference cell filled with the same volume of buffer, in the temperature range 5-90 °C, at a rate of 0.5 °C/ min. Repeated buffer versus buffer scans were carried out with 10 mM sodium cacodylate containing 10 mM NaCl to obtain an appropriate reproducible baseline that was then subtracted from the sample versus buffer scan. The average sample scans were normalized for concentration, and the progressive baseline mode available in the analysis package (Origin 7. 0) was used to integrate the resulting curve, ∆Cp dT, to yield the molar unfolding calorific enthalpy, ∆Hcal, near the melting temperature. The corresponding calorimetrically derived van’t Hoff transition enthalpy (∆HVH) was also determined by analyzing the transition width of the heat capacity curves, while the midpoint of the transition gave the melting temperature Tm. The difference between pre- and postdenaturation baseline extrapolated to the midpoint of the transition extracts the heat capacity change, ∆Cp, associated with duplex melting.22 The obtained ∆Cp values were then used to furnish a complete thermodynamic profile at 25 °C, using the following equations:23
∆H(T) ) ∆Hcal + ∆Cp(T-Tm)
(7)
T∆S(T) ) ∆Hcal(T/Tm) + ∆CpT ln(T/Tm)
(8)
∆G(T) ) ∆Hcal[(1-T/Tm) + ∆Cp(T-Tm- ln(T/Tm)]
(9) where T ) 25 °C and ∆Hcal represents the enthalpy change around the melting temperature. Furthermore, the effects of TANA incorporation on the thermodynamic parameters can be represented as the difference in the ∆H and ∆S for hybridization between the TANA-substituted duplex and the unmodified
Energetics of TANA-modified Duplexes
J. Phys. Chem. B, Vol. 113, No. 9, 2009 2947 where dR/dt is the rate of change of the SPR response signal, R and Rmax are the measured and maximum reponse signal measured with binding, C is the analyte concentration, and ka and kd are the association and dissociation rates, respectively. The binding constant, KA is calculated as ka/kd. At equilibrium, dR/dt ) 0, and eq 1 can be written as
Req ) KARmax-KAReq C
Figure 1. Circular dichroism spectra DNA duplexes containing varying number of TANA modificationssTANA-0 (0), TANA-1 (O), TANA-2 (∆), and TANA-3 (3) modified duplexes (5 um)shave been depicted at 10 °C in 10 mM sodium cacodylate (pH 7.4) containing 10 mM NaCl.
duplex (∆∆H ) ∆HTANA/DNA•DNA - ∆HDNA•DNA and ∆∆S ) ∆SLNA+ DNA•DNA - ∆SDNA•DNA). Surface Plasmon Resonance Study (SPR). SPR measurements were performed with a Pharmacia BIAcore 2000 system and streptavidin-coated sensor chips (Sensor chip SA, BIAcore). The streptavidin sensor surface injections of 1 M NaCl in 50 mM NaOH were followed by extensive washing with buffer. The 5′ biotinylated oligomers (SBS Genetech) were dissolved in running buffer (filtered and degassed, 10 mM HEPES buffer (pH 7.4), 100 mM NaCl, and 0.005% surfactant P20 from BIAcore) before noncovalent immobilization on flow cell 2 to attain a binding of ∼140 response units (RU). Flow cell 1 was left blank as a control to account for any signal generated due to bulk solvent effect or any other effect not specific to the DNA/ TANA interaction, which was subtracted from the signal obtained in flow cell 2. Manual injection at the rate of 2 µL/ min was used to achieve long contact times and to control the amount of bound DNA on the surface. The DNA-immobilized surface was exposed to the running buffer for 2 h at a flow rate of 2 µL/min before the modified complementary strands were added. Samples of TANA were prepared fresh in running buffer by serial dilutions from a stock solution immediately before the experiment. All procedures were carried out at 15 °C using repetitive cycles of sample injection and regeneration. TANA substituted oligonucleotides at different concentrations ranging from 0.0625 to 2 µM prepared in the running buffer and were injected (at 20 µL min-1 for 300 s) in random series to avoid any systematic error. Running buffer was not sufficient for dissociation of the hybridized complementary strands. Hence, surfaces were regenerated after hybridization by passing 1 M NaCl, 50 mM NaOH solution at the flow rate of 20 µL/min for 1 min. All data were obtained in duplicates. The reference response from the blank cell was subtracted from the response in each cell containing DNA to give a signal (RU, response units) that is directly proportional to the amount of bound DNA strand. Sensorgrams, RU versus time, at different concentrations for binding of each modified strand to DNA were obtained, and the RU in the steady-state region was determined by linear averaging over a selected time span. Data was analyzed using BIAevaluation 3.1.1. Under pseudo first-order conditions, where the free analyte concentration is held constant in the flow cell, the binding is described by:
dR ) kaC(Rmax-R)-kdR dt
(10)
Req )
RmaxKAC (1 + KAC)
(11)
(12)
Req is the measured response at equilibrium, and values of Req are obtained at a series of injected analyte C concentrations. The steady state response when plotted versus analyte concentrations and fit to Langmuir isotherm for a molecular interaction provides the binding affinity (KA) of immobilized molecule to its target and Rmax.24 Results Circular Dichroism Study. The CD spectra for all duplexes at 10 °C and at duplex concentration 5 µM, are shown in Figure 1. As seen in the figure, the unmodified duplexes displayed a positive peak at 275 nm and a negative peak at 250 nm, which is the characteristic of a helix in a B-conformation. The CD spectra of modified duplexes, however, differed reasonably from the unmodified B-form duplex, and showed significant shift in
Figure 2. UV melting curves representing the effect of substitution on the thermostability of the modified duplexes. Individual traces represent the TANA-0 (0), TANA-1 (O), TANA-2 (∆) and TANA-3 (3) modified duplexes at 5 uM concentration in 10 mM sodium cacodylate buffer containing 100 mM NaCl.
Figure 3. Tm dependence on duplex concentration is shown for TANA-0 (0), TANA-1 (O), TANA-2 (∆), and TANA-3 (3) modified duplexes in 10 mM sodium cacodylate buffer containing 10 mM NaCl
2948 J. Phys. Chem. B, Vol. 113, No. 9, 2009
Kaur et al.
TABLE 2: Thermodynamic Parameters Obtained by UV for Helix-Coil Transition duplex
Tm °C at 110 mM Na+
Tm decline per modification
∆HMATH (kcal/mol)
∆HvH (kcal/mol)
∆S e.u.
∆G(25 °C) kcal/mol
TANA-0 TANA-1 TANA- 2 TANA- 3
41.0 ((0.8) 38.4 ((0.7) 31.3 ((0.7) 27.8 ((0.5)
2.6 4.8 4.4
-74.5 ((2.6) -71.0 ((2.1) -66.6 ((1.9) -63.7 ((1.9)
-80.6 ((7.2) -68.0 ((5.4) -63.3 ((3.7) -60.5 ((3.6)
-229 ((25) -191 ((19) -179 ((14) -173 ((14)
-12.2 ((0.2) -11.0 ((0.2) -9.9 ((0.4) -8.8 ((0.5)
the positive and negative peaks. The spectra in this case exhibited a positive peak at 270 nm, whereas the negative peak shifted toward 242 nm with a significant decrease in the amplitudesa signature characteristic of A-type helices. The shift observed for both positive and negative peaks increased with the increase in the number of TANA-substitutions in the modified duplex. This change in signature observed for the modified duplexes with respect to the unmodified duplex suggests that that introduction of the TANA-linkages causes a change in the global conformation of the duplexsfrom a B-type toward A-type conformation of the helix.25 UV Melting Study. The UV-melting curves for duplexes at 260 nm and the Tm dependence on concentration have been shown in Figure 2 and 3, respectively. Interestingly, all the duplexes exhibited a continuous thermal activity in the premelting domain, as is reflected by the strongly sloped lower baselines (Figure 2). Figure 2 also reflects that an increase in the number of TANA modifications led to a decrease in the thermal stability of the duplex. The decline in Tm per modification at 5 µM duplex concentration and at 100 mM NaCl was found to be 3.3 (TANA-1), 5.0 (TANA-2), and 4.9 °C (TANA3), as shown in Table 2. For a given modification an increase in duplex concentration from 2-40 µM led to a concomitant rise in Tm. The thermodynamic parameters obtained from 1/Tm versus ln CT/4 plots of the respective modifications are listed in Table 2. Thermodynamic parameters were also derived by curve fitting in mathemetica 5.1 and correlated well with our experimental data within the experimental error limit. Inspection of the thermodynamic data obtained by both methods reveals that an increase in number of modification in the native duplex lowers the thermodynamic stability (∆G) of the duplex by making the ∆HvH of the system less negative. The same trend could be observed for the mathematica-derived enthalpy change, ∆HMATH (Table 2). DSC Melting Study. The model-independent thermodynamic parameters and the associated heat capacity functions accompanying helix-coil transitions were observed using DSC.26 Typical DSC melting curves for individual duplexes are
Figure 4. Typical DSC curves of duplex melting in 10 mM sodium cacodylate buffer (pH 7.4), 10 mM NaCl for TANA-0 (0), TANA-1 (O), TANA-2 (∆), and TANA-3 (3) modified duplexes. The inset describes estimation of ∆Cp by extrapolation of pre- and postdenaturation baselines to midpoint of transition.
represented in Figure 4. As seen from the figure, predenaturation baselines exhibited a deviation from linearity, whereas the postdenaturing baselines were nearly flat. The difference between predenaturation baseline (drawn as a tangent) and postdenaturation baseline (extended horizontally) when extrapolated to the midpoint of the transition, as shown in Figure 4 inset, gave the heat capacity change, ∆Cp.26,27 The so-obtained ∆Cp values accompanying each duplex formation has been tabulated in Table 3. The average ∆Cp change observed per duplex was found to be 832.5 cal mol-1 K-1, yielding an averaged base pair change of 59.5 cal (mol of base pair)-1 K-1. Determination of ∆Cp allows extrapolation thermodynamic parameters at any given temperature, T, using eqs 6-8. We have used this set of equations to obtain complete thermodynamic profile for individual duplex formation at 25 °C (Table 3). Duplex formation is known to be accompanied by favorable enthalpy (negative ∆H) and an unfavorable entropy (a negative ∆S) change. This is also reflected in our tabulated values, presented in Table 3. The influence of TANA substitution on thermodynamic parameters, with respect to the unmodified duplex, is also depicted (∆∆H and ∆∆S). It appears that introduction of an extra modification in the DNA duplex makes the enthalpy term less favorable, which overrides the effect of a decrease in unfavorable entropy parameter. The same trend could also be observed for the DSC-derived van’t Hoff enthalpy change; however, the observed values for ∆HVH were slightly less than that observed for ∆Hcal (∆Hcal/∆HVH > 1). Counterions Uptake upon Duplex Formation. Sodium ions are known to bind to DNA in various forms. These condensed ions contribute to the stability of helical structures by reducing the repulsive forces between phosphate groups along the chain. When the DNA duplex melts, there is a release of Na+ ions due to the reduction of charge density along the chain in conversion of helical duplex DNA to single strands. To estimate the effect of TANA linkages on the counterions uptake associated with the process of duplex formation, UV melting curves of duplexes (5 µM) at several salt concentrations (20-120 mM Na+) were studied. The increase in the salt concentration resulted in the shift of the melting curves to higher temperatures. This is the characteristic effect of salt, stabilizing the duplex state with a higher negative charge density parameter. The dependence of Tm on salt concentration is shown in Figure 5 for each duplex. From a linear regression analysis of the Tm vs ln[Na+] plots, slopes of melting of these duplexes in varying NaCl concentration have been calculated (Table 4). It has been shown that the values of these slopes have been proportional to the difference in the number of bound counterions in the single-stranded and the double-stranded states.28 The counterion uptake for each duplex (Table 4) over this range of salt concentration was calculated using eq 6. As shown in the table, an increase in modification led to a concomitant increase in the slope values; however, the observed change in the counterion uptake (∆nNa+) associated with increasing the number of modifications was under experimental error and therefore, insignificant.
Energetics of TANA-modified Duplexes
J. Phys. Chem. B, Vol. 113, No. 9, 2009 2949
TABLE 3: Thermodynamic Parameters Obtained by DSC for the Formation of Duplexes duplex
Tm °C
TANA-0 TANA-1 TANA- 2 TANA- 3
48.3 ((0.5) 45.4 ((0.5) 42.5 ((0.5) 37.7 ((0.5)
∆Hcal kcal/mol ∆Cp cal/mol · K -94.3 ((2.8) -80.7 ((2.2) -70.5 ((1.7) -65.0 ((1.6)
670 ((127) 890 ((133) 940 ((141) 830 ((149)
∆H(25 °C) kcal/mol
∆S(25 °C) e.u.
∆G(25 °C) kcal/mol
-109.9 ((0.16) -98.8 ((0.51) -86.9 ((0.88) -75.5 ((0.3)
-342 ((0.5) -310 ((1.6) -275 ((2.7) -243 ((1.0)
-8.0 ((0.01) -6.4 ((0.3) -5.0 ((0.7) -3.0 ((0.05)
∆∆H(25 °C) ∆∆G(25 °C) kcal/mol ∆∆S(25 °C) e.u. kcal/mol 11.0 22.9 34.3
31.6 66.4 94.9
1.6 3.1 4.8
TABLE 4: Thermodynamic Uptake of Counterions upon Formation of Duplexa duplex
∆Tm/∆Na+ (slopes)
∆nNa+ (per duplex)
∆∆nNa+
TANA-0 TANA-1 TANA- 2 TANA- 3
5.5 ((0.26) 6.3 ((0.31) 6.4 ((0.31) 6.6 ((0.33)
-2.42 ((0.25) -2.35 ((0.23) -2.34 ((0.23) -2.35 ((0.23)
0.07 0.08 0.07
a
Experiments were conducted at duplex concentrations 5 µM in 10 mM sodium cacodylate buffer containing 10 mM salt, adjusted to five different concentrations of sodium chloride. ∆nNa+ were derieved using equation ∆nNa+ ) 1.11(∆H/[RTm2])[∆Tm/∆(ln[Na+])].
Surface Plasmon Resonance. Typical sensorgrams for the binding of varying concentrations of unmodified and TANAmodified oligonucleotides to the immobilized cDNA strand are shown in Figure 6. For all the investigated duplexes, hybridization experiments were carried out using five different concentrations of the associating strand. The fit of the sensorgram using BIAevaluation software gave the association and dissociation rates constants for the interaction of unmodified and TANAmodified oligonucleotides with the immobilized complementary strand (Table 5). The dissociation rate constants (kd) reported here are the averages of values determined at different concentrations of the incoming strand. Req, the measured response unit at equilibrium, when plotted versus complementary strand concentrations and fit to Langmuir isotherm for molecular interactions provides the binding affinity (Figure 6c) and the maximum response at saturation of binding sites, which collectively reflects the amount of duplex formed. As seen from the table, the presence of TANA-linkages in the duplex leads to a moderate decrease in the association rate constant ka and a marked increase in the dissociation rate constant kd as compared to the unmodified duplexes. Relative to the unmodified duplex, the increase in kd for the triple-modified duplex was almost 5-fold whereas the decrease in ka was only about 2-fold. This results in a lower KA for the DNA/TANA · DNA association than for DNA · DNA association. Discussion Modification in the phosphodiester backbone of DNA has long been sought as an alternative to improve the antisense
Figure 6. Representive sensorgrams for the hybridization of DNA with complementary (a) TANA-0 and (b) TANA-3 oligonucleotides at five different concentrations ranging from 1.25 to 2 uM (top to bottom) at 15 °C in 10 mM HEPES pH 7.4, containing 0.005% surfactant and 100 mM NaCl. Panel c shows the binding affinity curves for TANA-0 (0), TANA-1 (O), TANA-2 (∆), and TANA-3 (3) modified duplexes.
TABLE 5: Effect of TANA-substitution on Dissociation (kd) and Association (ka) Rate Constants, and the (Calculated) Association Equilibrium Constants (KA) for Duplexes at 15 °Ca duplex
ka M-1 s-1
kd s-1
KA M-1
KAunmod/KAmod
TANA-0 TANA-1 TANA- 2 TANA- 3
1.2 ((0.2) × 103 0.8 ((0.1) × 103 0.6 ((0.1) × 103 0.6 ((0.1) × 103
1.2 ((0.2) × 10-4 2.0 ((0.3) × 10-4 3.0 ((0.4) × 10-4 6.0 ((0.9) × 10-4
1.0 ((0.1) × 107 4.0 ((0.6) × 106 2.0 ((0.3) × 106 1.0 ((0.1) × 106
2.5 10.5 10
a Experiments were done at 10 °C in 10 mM HEPES buffer (pH 7.4) with 100 mM NaCl.
Figure 5. Tm dependence on salt concentration is shown for TANA-0 (0), TANA-1 (O), TANA-2 (∆) and TANA-3 (3) modified duplexes.
efficacy of an oligonucleotide. Among the different kinds of modifications tested, backbone substitution with an amidelinkage has proven to be an obvious choice. This is ascribed to its increased stability of amide group in physiological conditions, its ability to penetrate in cells, its compatibility with conditions
2950 J. Phys. Chem. B, Vol. 113, No. 9, 2009 SCHEME 2: A representative of the compact N-type and the extended S-type conformation of TANA -linker
for solid phase synthesis, and finally, its achiral nature that circumvents the problem of the mixing of diastereomeric forms, during synthesis. The current study investigates the performance of TANA-modified oligonucleotides with DNA as a hybridizing partner. This recently developed linker consists of a five-atom amide link instead of a four-atom phosphate internucleoside linkage. As compared to the other amide-linked modifications, the synthesis of TANA dimer blocks is straightforward;8 however, as is true for other amide linkages, the possibility of partial double bond character (resonance effect) between the amide nitrogen and the carbonyl carbon of the TANA-linkage may hinder free rotation around this bond. The extended 5-atom backbone of TANA-linker, however, compensates for the reduced conformational flexibility of this amide bond, relative to the native phosphodiester backbone. Structural characterization using CD spectroscopy suggested that presence of modification produces an apparent change in the helical geometry from a B-type toward an A-type conformation. We attribute this to structural transition arising because of the lengthening of the linker chain connecting two consecutive sugar moieties in the nucleic acid. In contrast to a natural nucleic acid that contains a four-atom phoshodiester linker, TANA possesses a five-atom linker. To compensate for this extra atom between the sugars, the linker-backbone possibly folds into a much more compact 3′-endo or N-type sugar conformation, as depicted in Scheme 2. Unlike, a phosphodiester backbone where electrostatic repulsions between the phosphates maintain the chain in an extended conformation, the absence of charge in TANA-linkage allows more conformational flexibility, thereby increasing the feasibility of backbone folding. The shifting of sugar conformation toward an N-type sugar pucker (found in RNA), as against the S-type pucker prevalent in DNA, in turn shifts the helical geometry from a B-type to A-type. Such changes in sugar conformation per se affect the hydrogen bonding, the base-stacking interaction, and the hydration of the major/minor grooves, thereby considerably altering the binding efficiency of the oligonucleotide toward the complementary strand. Understanding the thermodynamic basis of duplex formation involves accounting for a number of physio-chemical factors such as base-pairing, base-pair stacking of the duplex at the low temperature, base-pairing stacking of the single strands at high temperatures, and the difference between counterion and
Kaur et al. hydration of the duplex versus single-stranded state. Using both spectroscopic and calorimetric techniques we observed that the presence of modification decreased the thermostability of the DNA duplex. However, a notable but not unusual observation was the appearance of continuous thermal activity before the onset of global melting domain. Such premelting is expected to arise from changes in hydration state of the duplex, which ultimately manifests as changes in duplex helical geometry and interbase stacking interactions.29-31 Further analysis of thermodynamic parameters obtained at 10 °C suggested that TANAinduced destabilization originates from a decrease in the favorable enthalpy term that overrides the effect of decrease in the unfavorable entropy. Although this unfavorable entropy contribution stems from the reduction in the number of molecules due to bimolecular association of two strands, and the uptake of counterions and water molecules, the favorable enthalpy parameter originates from the exothermic contributions of base pairing and base-pair stacking interactions. The fact that the presence of TANA produces a larger decrease in favorable enthalpy than in the unfavorable entropy term (Table 3) suggests that TANA-associated destabilization is largely enthalpic in nature, arising merely from the poor base-stacking interactions. A comparison of model dependent van’t Hoff enthalpy (∆HVH) and model independent calorific enthalpy (∆Hcal) reflected the nature of helix-to-coil transition. If the process is two-state, the ratio of ∆HVH/∆Hcal is equal to unity.32 In our case, the ∆HVH/ ∆Hcal ratio deviates from unity, suggesting that helix-to-coil transition in this study are not strictly two-state and might involve intermediates. Our observation is in concordance with the previous literature studies that suggest that two-state helixto-coil transitions are typically the trait of canonical duplexes containing eight or fewer bases.33-37 For longer duplexes, the calorific and the van’t Hoff enthalpies may be significantly different.35 Consequently, the van’t Hoff analysis of stability characteristics of these duplexes might lead to errors in estimation of thermodynamic parameters. Characterization of thermodynamic parameters in this case would entirely rely on calorific melting data. Calorimetric analysis allows direct extraction of the thermodynamic parameters accompanying duplex unfolding, these parameters are determined near the melting temperature, Tm, only. Since biological application of modified oligonucleotides would involve extrapolation of thermodynamic parameters to the relevant experimental temperature, a prior knowledge of heat capacity change (∆Cp) associated with helix-to-coil transition is required to account for the temperature dependence of ∆H and ∆S.22,23,38-41 Usually, these Cp changes originate from solvent effects that accompany burial of hydrophobic group and/ or aromatic surfaces upon folding.39 In our case we observed a positive ∆Cp accompanying the helix to coil transitions of these modified duplexes.26,27 The average value of ∆Cp was found to be 832.5 cal mol-1 K-1, which is equivalent to 59.5 cal (mol of base pairs)-1 K-1. Our ∆Cp values are thus in qualitative agreement with the literature values of 64 cal (mol of base pairs)-1 K-1 for long polymeric nucleic acids22,23 and an average value of 36 cal (mol of base pairs)-1 K-1 observed for small duplexes. We have used this value to deduce complete thermodynamic parameters at 25 °C (Table 3). To further characterize the thermodynamics we investigated the difference in the number of uptake of counterions (Na+) associated with the formation of modified-duplex and compared to that of unmodified DNA duplex. Although the uptake has little effect on the enthalpy change, it contributes significantly to the unfavorable entropy parameter of duplex formation. When
Energetics of TANA-modified Duplexes monitoring the Tm dependence of duplexes on salt concentration, it appears that the extent of salt stabilization is higher for modified duplexes, which is reflected in their higher slope values (∆Tm/∆(ln[Na+]). In contrast, the net counterions (∆nNa+) uptake associated with formation of individual duplex showed an insignificant change with increase in modification. This might be explained on the basis that incorporation of a TANA unit decreases the overall negative charge of the duplex, but at the same time it shifts the helical geometry to a form (A-type) that is known to posses a higher charge-density parameter.21 The interplay between these two oppositely acting factors thus accounts for the insignificant change and the absence of trend in the ∆nNa+ values. Once structural and thermodynamic evaluation have been completed, the next vital aspect to judge the functional performance of a modified oligonucleotide involves a detailed analysis of its hybridization kinetics to the complementary-target strand. Assessment of kinetic parameters particularly becomes important when different nucleotide chemistries exhibiting comparable thermodynamic stability with the target strand are compared. For TANA-substituted oligonucleotides an increase in number of substitutions led to a subsequent decrease in binding affinity to the complementary strand. The modified oligonucleotide binds to the complementary strand with the binding affinity order of 106, which was 10-fold lower than that observed for the binding of unmodified DNA strands. A detailed kinetic analysis undertaken to account for the limited stability of TANA-modified duplexes suggests that the origin of destabilization mainly stems from the higher rate dissociation contstants for the modified oligonucleotides as compared to the unmodified reference oligonucleotides. Conclusion This study provides a detailed insight into the effect of incorporation of TANA on hybridization thermodynamics, binding kinetics and counterion uptake associated with DNA duplex formation. We observed that backbone extension by fiveatom TANA-linker destabilized DNA duplexes. We conclude that this decreased stability mainly emanates from the increased dissociation rates constants of the modified duplexes, which in turn decrease the hybridization efficiency of the complementary strands. Thermodynamically, the presence of extra TANA substitution produces a larger decrease in the favorable enthalpy term that overrides the effect of lowering of unfavorable entropy term, thereby, making the overall process of duplex formation energetically less favorable than the unmodified duplex. Analysis of our DSC data allowed estimation of ∆Cp change accompanying each duplex formation, which could be used to extrapolate thermodynamic parameters at any given experimental temperature condition. Comparison between calorimetric and van’t Hoff enthalpies revealed helix-to-coil transitions for both modified and unmodified duplexes associated with nontwo-state transitions. Furthermore, optical melting data showed a modest change in the counterion uptake by the modified duplexes, attributed to the modifications induced in the charge and the conformation of the helix. Overall, our data shows that TANA modification destabilizes the DNA helix. The use of such modification, either alone or in combination with other modifiednucleotide-chemistries, would thus allow for optimization of stability characters of the probe.
J. Phys. Chem. B, Vol. 113, No. 9, 2009 2951 Acknowledgment. This work was supported by Council of Scientific and Industrial Research, India (Comparative Genomics and Biology of noncoding RNA). K.G. thanks CSIR for fellowship. P.S. was financially supported by the Department of Science and Technology, New Delhi, India. References and Notes (1) Kurreck, J. Eur. J. Biochem. 2003, 270, 1628. (2) Crooke, S. T. Methods Enzymol. 2000, 313, 3. (3) Ratilainen, T.; Holmen, A.; Tuite, E.; Haaima, G.; Christensen, L.; Nielsen, P. E.; Norden, B. Biochemistry 1998, 37, 12331. (4) Freier, S. M.; Altmann, K. H. Nucleic Acids Res. 1997, 25, 4429. (5) Kaur, H.; Babu, B. R.; Maiti, S. Chem. ReV. 2007, 107, 4672. (6) Trempe, J. F.; Wilds, C. J.; Denisov, A. Y.; Pon, R. T.; Damha, M. J.; Gehring, K. J. Am. Chem. Soc. 2001, 123, 4896. (7) Walker, G. T. Nucleic Acid Research 1988, 16, 3091. (8) Gogoi, K.; Gunjal, A. D.; Phalgune, U. D.; Kumar, V. A. Org. Lett. 2007, 9, 2697. (9) Schon¨ing, K. U.; Scholz, P.; Wu, X.; Guntha, S.; Delgado, G.; Krishnamurthy, R.; Eschenmoser, A. Chim. Acta 2002, 85, 4111. (10) Reck, F.; Wippo, H.; Kudick, R.; Bolli, M.; Ceulemans, G.; Krishnamurthy, R.; Eschenmoser, A. Org. Lett. 1 1990, 10, 1531. (11) Zhang, L.; Peritz, A.; Meggers, E. J. Am. Chem. Soc. 2005, 127, 4174. (12) De Mesmaeker, A.; Waldner, A.; Wendeborn, S.; Wolf, R. M. Pure Appl. Chem. 1997, 3, 437. (13) Nina, M.; Fonne´-Pfister, R.; Beaudegnies, R.; Chekatt, H.; Jung, P. M.; Murphy-Kessabi, F.; De Mesmaeker, A.; Wendeborn, S. J. Am. Chem. Soc. 2005, 127, 6027. (14) De Mesmaeker, A.; Jouanno, C.; Wolf, R. M.; Wendeborn, S. Bioorg. Med. Chem. Lett. 1997, 7, 447. (15) De Napoli, L.; Iadonisi, A.; Montesarchio, D.; Varra, M.; Piccialli, G. Bioorg. Med. Chem. Lett. 1995, 5, 1647. (16) Wilds, C. J.; Minasov, G.; von Matt, P.; Altmann, K.-H.; Egli, M. Nucleosides, Nucleotides Nucleic Acids 2000, 20, 991. (17) Pallan, P. S.; von Matt, P.; Wilds, C. J.; Altmann, K. H.; Egli, M. Biochemistry 2006, 45, 8048. (18) Marky, L. A.; Blumenfeld, K. S.; Kozlowski, S.; Breslauer, K. J. Biopolymers. 1983, 22, 1247. (19) Marky, L. A.; Breslauer, K. J. Biopolymers. 1987, 26, 1601–1620. (20) Mergny, J. L.; Lacroix, L. Oligonucleotides 2003, 13, 515. (21) Rentzeperis, D.; Kharakoz, D. P.; Marky, L. A. Biochemistry. 1991, 30, 6276. (22) Chalikian, T. V.; Volker, J.; Plum, G. E.; Breslauer, K. J. Proc. Natl. Acad. Sci. U. S. A. 1999, 96, 7853. (23) Tikhomirova, A.; Beletskaya, I. V.; Chalikian, T. V. Biochemistry. 2006, 45, 10563. (24) O’Shannessy, D. J.; Brigham-Burke, M.; Soneson, K. K.; Hensley, P.; Brooks, I. Anal. Biochem. 1993, 212, 457. (25) Johnson, W. C. In Circular Dichroism: Principles and Applications; Nakanishi, K., Berova, N., Woody, R. W., Eds.; VCH: New York, 1994, 523. (26) Rouzina, I.; Bloomfield, V. A. Biophys. J. 1999, 77, 3242. (27) Tikhomirova, A.; Taulier, N.; Chalikian, T. V. J. Am. Chem. Soc. 2004, 126, 16387. (28) Record, T. M., Jr.; Anderson, C. F.; Lohman, T. M. Q. ReV. Biophys. 1978, 11, 103. (29) Palecek, E. Prog. Nucleic Acid Res. Mol. Biol. 1976, 18, 151. (30) Erfurth, S. C.; Peticolas, W. L. Biopolymers 1975, 15, 247. (31) Herrera, J. E.; Chaires, J. B. Biochemistry 1989, 28, 1993. (32) Kaya, H.; Chan, H. S. Proteins 2000, 40, 637. (33) Plum, G. E.; Grollman, A. P.; Johnson, F.; Breslauer, K. J. Biochemistry 1995, 34, 16148. (34) Holbrook, J. A.; Capp, M. W.; Saecker, R. M.; Record, M. T., Jr. Biochemistry 1999, 38, 8409. (35) Breslauer, K. J.; Frank, R.; Blocker, H.; Marky, L. A. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 3746. (36) Kaur, H.; Wengel, J.; Maiti, S. Biochemistry 2008, 47, 1218. (37) Kaur, H.; Arora, A.; Wengel, J.; Maiti, S. Biochemistry 2006, 45, 7347. (38) Holbrook, J. A.; Capp, M. W.; Saecker, R. M.; Record, M. T. Biochemistry 1999, 38, 8409. (39) Mikulecky, P. J.; Feig, A. L. Biopolymers 2006, 82, 38. (40) Breslauer, K. J.; Freire, E.; Straume, M. Methods Enzymol. 1992, 211, 533. (41) Mikulecky, P. J.; Feig, A. L. Biochemistry 2006, 45, 604.
JP808747G
