Nearest neighbor parameters for Watson–Crick complementary heteroduple
http://www.100md.com
《核酸研究医学期刊》
1 Department of Chemistry, University of Rochester RC Box 270216, Rochester, NY 14627, USA 2 Institute of Bioorganic Chemistry, Polish Academy of Sciences 60-714 Poznan, Noskowskiego 12/14, Poland 3 Center for Pediatric Biomedical Research, University of Rochester School of Medicine and Dentistry Rochester, NY 14642, USA 4 Department of Biochemistry and Biophysics, University of Rochester School of Medicine and Dentistry Rochester, NY 14642, USA
*To whom correspondence should be addressed. Tel: +1 585 275 3207; Fax: +1 585 276 0205; Email: turner@chem.rochester.edu
ABSTRACT
Results from optical melting studies of Watson–Crick complementary heteroduplexes formed between 2'-O-methyl RNA and RNA oligonucleotides are used to determine nearest neighbor thermodynamic parameters for predicting the stabilities of such duplexes. The results are consistent with the physical model assumed by the individual nearest neighbor-hydrogen bonding model, which contains terms for helix initiation, base pair stacking and base pair composition. The sequence dependence is similar to that for Watson–Crick complementary RNA/RNA duplexes, which suggests that the sequence dependence may also be similar to that for other backbones that favor A-form RNA conformations.
INTRODUCTION
Oligonucleotides are used for many applications, ranging from diagnostics (1–4) to therapeutics (5–9) to nanotechnology (10,11). The thermodynamics of nucleic acid duplex formation facilitates rational design of sequences for the various applications (12–14). The thermodynamics of duplex formation is dependent on the backbone of the nucleic acid. For example, the sequence dependence of the thermodynamics of DNA/DNA (15,16), RNA/RNA (17,18) and DNA/RNA (19) duplexes differ. All, however, can be approximated well by nearest neighbor models when only Watson–Crick base pairs are formed. Thus, it is relatively easy to predict the thermodynamics of Watson–Crick paired duplexes from sequence (16,18–22). Here, optical melting studies are analyzed to provide nearest neighbor thermodynamic parameters for formation of 2'-O-methyl RNA/RNA duplexes that are Watson–Crick complementary. The 2'-O-methyl RNA and other 2'-O-alkyl backbones are particularly useful for hybridization to RNA because they favor A-form helical structure and are more resistant than RNA or DNA to nuclease digestion (23–26).
MATERIALS AND METHODS
Experimental
Synthesis and purification of oligonucleotides was done as previously described (27). The buffer for melting experiments was 100 mM NaCl, 20 mM sodium cacodylate and 0.5 mM Na2EDTA, pH 7.0. Oligonucleotide single strand concentrations were determined from absorbances >80°C with extinction coefficients approximated by a nearest neighbor model (28,29). The sequence dependence of extinction coefficients for 2'-O-methyl and RNA strands was assumed to be identical. Melting curves were measured at 260 nm with a heating rate of 1°C/min from 0 to 90°C on a Beckman DU640 spectrophotometer with a water cooled Peltier thermoprogrammer. Melting curves were analyzed and thermodynamic parameters were calculated on the basis of a two-state model with the program MeltWin 3.5 (30). With one exception, agreement within 15% was found for thermodynamic parameters calculated from averaging parameters derived from the shapes of melting curves and from the following equation (20):
(1)
This agreement is consistent with the two-state model.
Parameter fitting
The measured thermodynamic parameters were fit to the individual nearest neighbor-hydrogen bonding (INN-HB) model (18) by multiple linear regression with the program Analyse-It v.1.71 (Analyse-It Software, Ltd, Leeds, England; www.analyse-it.com), which expands Microsoft Excel. Only duplexes that melted in a two-state manner were included in the fit. Measured parameters from versus ln(CT/4) plots were used as the data for the calculations. Error limits reported for the experimental data reflect the scatter in 's when fit to Equation 1. Systematic errors are typically larger, however, and difficult to estimate (18). For example, the melting is not truly two-state because the stacking in the single strand conformations is dependent on temperature and sequence. Therefore, all duplexes included in the fit were given equal weight.
RESULTS
Table 1 lists measured thermodynamic parameters for several 2'-O-methyl RNA/RNA duplexes. Only the duplex m(5'-CGAAGUGAA)/r(3'-GCUUCACUU) does not melt in an apparent two-state manner as revealed by a >15% difference between the H°s derived from averaging fits to the shapes of melting curves and from the versus ln (CT/4) plot. With the exception of m(5'-CGAAGUGAA)/r(3'-GCUUCACUU), results in Table 1 were combined with previously reported results (27) listed in Table 2 and fit to the INN-HB nearest neighbor model (18) to give the nearest neighbor thermodynamic parameters listed in Table 3.
Table 1 Thermodynamic parameters of heteroduplex formation between 2'-O-methyl RNA and oligoribonucleotides in 0.1 M NaCl, pH 7a
Table 2 Measured and predicted (in parenthesis) G°37, H°, S° and the percentage difference between measured and predicted values for Watson–Crick complementary 2'-O-methyl RNA/RNA duplexes in 0.1 M NaCl, pH 7
Table 3 Thermodynamic parameters for INN-HB nearest neighbor model applied to 2'-O-methyl RNA/RNA heteroduplexes in 0.1 M NaCl, pH 7
On a percentage basis, the errors in nearest neighbor parameters for G°37 are much smaller than for H° and S°. This is expected from the high correlation of errors in H° and S°, typically with R2 > 0.99 (18,31,32). While the individual errors in nearest neighbor parameters for H° and S° are large, the percentage errors in predicting H° and S° for duplex formation are smaller. This is because the values of H° and S° are given by the sums of the nearest neighbors, but the errors propagate as the square root of the sum of the squares of the errors. Table 2 lists the predicted values of G°37, H° and S° and their percentage differences from measured values for the oligonucleotides studied. The range of percentage differences is 0–15, 0.04–23 and 0.1–26% for G°37, H° and S°, respectively, while the average differences are, respectively, 2.4, 6.7 and 7.7%. The worst percentage prediction of G°37 differs from the measured value by 0.73 kcal/mol, which translates to a difference of 3-fold in measured and predicted association constants. The thermodynamics for the duplex m(5'-CGAAGUGAA)/r(3'-GCUUCACUU) (Table 1) are also predicted reasonably well even though it does not melt in a two-state manner and was not used in fitting the nearest neighbor parameters. The predicted values for G°37, H° and S° are –9.87 kcal/mol, –78.3 kcal/mol and –220.6 eu while the measured values are –10.56 kcal/mol, –96.6 kcal/mol and –277.6 eu, respectively.
DISCUSSION
Thermodynamic parameters for nucleic acid duplexes are useful for designing sequences for many applications, including diagnostics, therapeutics and nanotechnology (12–14,33). The 2'-O-methyl backbone and other 2'-O-alkyl backbones are particularly useful for binding to RNA because they favor A-form helixes and are chemically stable relative to DNA and RNA backbones. Thus for example, 2'-O-alkyl backbones have been used to flank a ‘gap’ DNA sequence in order to decrease nuclease digestion of the oligonucleotide while providing a long enough pairing between DNA and RNA to induce RNase H to cleave an RNA target (34). Reduction of gene expression by an RNA interference mechanism has been demonstrated with siRNA duplexes having completely 2'-O-methyl modified sense strands (35). The 2'-O- methyl modification is also used in aptamers (7), including the commercially successful therapeutic, Macugen (8).
There are several ways to analyze the data in Table 2. In principle, it is possible to fit the data to 20 parameters each for G°37, H° and S° (21,22). We chose, however, to fit the data to the 18 parameters of the INN-HB model, which ascribes separate parameters to the 16 different nearest neighbor stacks, to the average difference between a CG and UA pair, and to duplex initiation (18). This simplifies the 20 parameter model by assuming that the parameters for initiation are independent of the nature of the terminal base pairs and that terminal base pairs are equivalent to internal base pairs. With the 18 parameter fit, the highest P-values for G°37 were 0.0007 and 0.0005 for the terminal AU and m(5'-AA)/r(3'-UU) parameters, respectively. All other P-values were <0.0001. For H°, however, P-values were 0.29, 0.15 and 0.12 for the initiation, m(5'-GA)/r(3'-CU), and m(5'-CA)/r(3'-GU) parameters, respectively. After correction for salt dependence, the parameters in Table 3 can be coupled with the existing parameters for RNA loops and incorporated into dynamic programming algorithms for predicting and designing secondary structures (14,17,36–38) and can be directly compared with the 12 INN-HB parameters for RNA (18).
A comparison of the G°37 parameters for 2'-O-methyl RNA/RNA duplexes at 0.1 M NaCl in Table 3 with G°37 parameters determined for RNA/RNA duplexes at 1 M NaCl (18) shows that on average the RNA/RNA parameters for G°37 of base pair stacks are more favorable by a value of 0.26 kcal/mol (Table 3). This is likely because of higher salt concentration. The sequence dependence of stability is similar for 2'-O-methyl RNA/RNA and RNA/RNA duplexes. For example, base pair stacks with two AU pairs are less stable than base pair stacks with one AU and one GC pair; the 5'-CG/3'-GC nearest neighbor has stability similar to nearest neighbors with one AU and one GC pair, but other stacks with two GC pairs are more stable. The parameters for terminal AU pairs are 0.30 and 0.45 kcal/mol for 2'-O-methyl RNA/RNA and RNA/RNA duplexes, respectively. In the INN-HB model, this parameter is used to account for the fact that two sequences can have the same nearest neighbor base pair stacks but differ by one in the number of AU and GC pairs (18). Thus it accounts for half of the difference in hydrogen bonding free energy between AU and GC pairs. The magnitude and sign of the 0.30 kcal/mol is consistent with this physical interpretation of the model. The similarity of the sequence dependence of parameters suggests that the programs for predicting RNA secondary structure (17,36–38) can approximate 2'-O-methyl modified nucleotides as unmodified RNA, at least in base paired regions. This is because the uncertainties in the sequence dependence of loop stabilities is probably larger than the difference in sequence dependence of RNA/RNA and 2'-O-methyl RNA/RNA base pairs. Moreover, natural RNA typically has relatively few 2'-O-methyl substitutions.
The similarity of the free energy parameters for 2'-O-methyl RNA/RNA and RNA/RNA duplexes also suggests that either set may be useful for predicting the sequence dependence of stabilities of base paired duplexes with other backbones that favor A-form RNA conformations. This would include other 2'-O-alkyl, 2'-fluoro and N3'–P5'phosphoramidate backbones (24,39,40) and chimeras that include a mixture of such backbones. For example, hammerhead ribozymes and aptamers have been designed with a mixture of RNA and 2'-fluoro backbones in order to increase resistance to nuclease digestion (8,41,42). The predictions would be rough approximations, but sufficient for many applications. As expected, 2'-O-methyl RNA/RNA duplexes appear more stable than DNA/RNA duplexes. The nearest neighbour stacking parameters for 2'-O-methyl RNA/RNA in 0.1 M NaCl average 0.4 kcal/mol more stable at 37°C than those for DNA/RNA duplexes in 1 M NaCl (19).
The average error limits for the 2'-O-methyl RNA/RNA nearest neighbor parameters for G°37 and H° are 0.17 and 3.7 kcal/mol, respectively (Table 3). These are roughly 2- and 5-fold larger than the average error limits for RNA/RNA (18) and DNA/DNA (16) parameters, respectively, where only 12 parameters can be fit and the number of duplexes is larger. Direct comparison with 17 and 18 parameter fits to data for DNA/RNA duplexes is not possible because error limits for individual parameters were not reported (19,22).
The error limits for the nearest neighbor parameters for H° and S° for 2'-O-methyl RNA/RNA are large. There are many potential reasons why the nearest neighbor model does not fit the experimental H° and S° values particularly well. The two-state model for fitting the melting curve data is an approximation (43–46). The high temperature state for a duplex transition is the single strand, but stacking in these single strands is temperature dependent, thereby producing a temperature dependent H° that is not included in the model. While the agreement between H°s determined by fitting melting curves and by plotting versus ln(CT/4) is consistent with two-state behavior over the temperature range of the transition, it is not sufficient to insure two-state behavior even over this range. In addition, the INN-HB nearest neighbor model is a simple approximation. The thermodynamics may depend on more than the stacking of nearest neighbors and number of hydrogen bonds. In general, parameters obtained from optical melting data are best near the melting temperatures of the oligonucleotides. The range of melting temperatures at 0.1 mM total strand concentration for the duplexes used to determine the nearest neighbor parameters range from 28 to 55°C with an average of 44°C. Thus the predicted G°s should be reasonable from 25 to 65°C.
ACKNOWLEDGEMENTS
This work was supported by Polish State Committee for Scientific Research (KBN) grant 2 PO4A03729 to R.K., NIH grant 1RO3 TW1068 to R.K. and D.H.T., and NIH grant GM22939 to D.H.T. A.C. is a recipient of a fellowship from the President of the Polish Academy of Sciences. Funding to pay the Open Access publication charges for this article was provided by NIH.
REFERENCES
Bodrossy, L. and Sessitsch, A. (2004) Oligonucleotide microarrays in microbial diagnostics Curr. Opin. Microbiol, . 7, 245–254 .
Heller, R.A., Schena, M., Chai, A., Shalon, D., Bedilion, T., Gilmore, J., Woolley, D.E., Davis, R.W. (1997) Discovery and analysis of inflammatory disease-related genes using cDNA microarrays Proc. Natl Acad. Sci. USA, 94, 2150–2155 .
Li, H.X. and Rothberg, L.J. (2004) Label-free colorimetric detection of specific sequences in genomic DNA amplified by the polymerase chain reaction J. Am. Chem. Soc, . 126, 10958–10961 .
Long, W.H., Xiao, H.S., Gu, X.M., Zhang, Q.H., Yang, H.J., Zhao, G.P., Liu, J.H. (2004) A universal microarray for detection of SARS coronavirus J. Virol. Methods, 121, 57–63 .
Zamecnik, P.C. and Stephenson, M.L. (1978) Inhibition of Rous-sarcoma virus-replication and cell transformation by a specific oligodeoxynucleotide Proc. Natl Acad. Sci. USA, 75, 280–284 .
Milligan, J.F., Matteucci, M.D., Martin, J.C. (1993) Current concepts in antisense drug design J. Med. Chem, . 36, 1923–1937 .
Burmeister, P.E., Lewis, S.D., Silva, R.F., Preiss, J.R., Horwitz, L.R., Pendergrast, P.S., McCauley, T.G., Kurz, J.C., Epstein, D.M., Wilson, C., et al. (2005) Direct in vitro selection of a 2'-O-methyl aptamer to VEGF Chem. Biol, . 12, 25–33 .
Ruckman, J., Green, L.S., Beeson, J., Waugh, S., Gillette, W.L., Henninger, D.D., Claesson-Welsh, L., Janji, N. (1998) 2'-fluoropyrimidine RNA-based aptamers to the 165-amino acid form of vascular endothelial growth factor (VEGF165) J. Biol. Chem, . 273, 20556–20567 .
Goodchild, J. (2004) Oligonucleotide therapeutics: 25 years agrowing Curr. Opin. Mol. Ther, . 6, 120–128 .
Chworos, A., Severcan, I., Koyfman, A.Y., Weinkam, P., Oroudjev, E., Hansma, H.G., Jaeger, L. (2004) Building programmable jigsaw puzzles with RNA Science, 306, 2068–2072 .
Seeman, N.C. (2005) From genes to machines: DNA nanomechanical devices Trends Biochem. Sci, . 30, 119–125 .
Matveeva, O.V., Mathews, D.H., Tsodikov, A.D., Shabalina, S.A., Gesteland, R.F., Atkins, J.F., Freier, S.M. (2003) Thermodynamic criteria for high hit rate antisense oligonucleotide design Nucleic Acids Res, . 31, 4989–4994 .
Schwarz, D.S., Hutvagner, G., Du, T., Xu, Z.S., Aronin, N., Zamore, P.D. (2003) Asymmetry in the assembly of the RNAi enzyme complex Cell, 115, 199–208 .
Dirks, R.M., Lin, M., Winfree, E., Pierce, N.A. (2004) Paradigms for computational nucleic acid design Nucleic Acids Res, . 32, 1392–1403 .
SantaLucia, J. and Hicks, D. (2004) The thermodynamics of DNA structural motifs Annu. Rev. Biophys. Biomolec. Struct, . 33, 415–440 .
SantaLucia, J., Jr. (1998) A unified view of polymer, dumbbell, and oligonucleotide DNA nearest-neighbor thermodynamics Proc. Natl Acad. Sci. USA, 95, 1460–1465 .
Mathews, D.H., Disney, M.D., Childs, J.L., Schroeder, S.J., Zuker, M., Turner, D.H. (2004) Incorporating chemical modification constraints into a dynamic programming algorithm for prediction of RNA secondary structure Proc. Natl Acad. USA, 101, 7287–7292 .
Xia, T.B., SantaLucia, J., Burkard, M.E., Kierzek, R., Schroeder, S.J., Jiao, X.Q., Cox, C., Turner, D.H. (1998) Thermodynamic parameters for an expanded nearest-neighbor model for formation of RNA duplexes with Watson–Crick base pairs Biochemistry, 37, 14719–14735 .
Sugimoto, N., Nakano, S., Katoh, M., Matsumura, A., Nakamuta, H., Ohmici, T., Yoneyama, M., Sasaki, M. (1995) Thermodynamic parameters to predict stability of RNA/DNA hybrid duplexes Biochemistry, 34, 11211–11216 .
Borer, P.N., Dengler, B., Tinoco, I., Uhlenbeck, O.C. (1974) Stability of ribonucleic-acid double-stranded helices J. Mol. Biol, . 86, 843–853 .
Gray, D.M. (1997) Derivation of nearest-neighbor properties from data on nucleic acid oligomers.1. Simple sets of independent sequences and the influence of absent nearest neighbors Biopolymers, 42, 783–793 .
Gray, D.M. (1997) Derivation of nearest-neighbor properties from data on nucleic acid oligomers.2. Thermodynamic parameters of DNA—RNA hybrids and DNA duplexes Biopolymers, 42, 795–810 .
Cramer, H. and Pfleiderer, W. (2000) Nucleotides LXIV: synthesis, hybridization and enzymatic degradation studies of 2'-O-methyl-oligoribonucleotides and 2'-O-methyl/deoxy gapmers Nucleosides Nucleotides Nucleic Acids, 19, 1765–1777 .
Lesnik, E.A. and Freier, S.M. (1998) What affects the effect of 2'-alkoxy modifications? 1. Stabilization effect of 2'-methoxy substitutions in uniformly modified DNA oligonucleotides Biochemistry, 37, 6991–6997 .
Majlessi, M., Nelson, N.C., Becker, M.M. (1998) Advantages of 2'-O-methyl oligoribonucleotide probes for detecting RNA targets Nucleic Acids Res, . 26, 2224–2229 .
Sproat, B.S., Lamond, A.I., Beijer, B., Neuner, P., Ryder, U. (1989) Highly efficient chemical synthesis of 2'-O-methyloligoribonucleotides and tetrabiotinylated derivatives—novel probes that are resistant to degradation by RNA or DNA specific nucleases Nucleic Acids Res, . 17, 3373–3386 .
Kierzek, E., Ciesielska, A., Pasternak, K., Mathews, D.H., Turner, D.H., Kierzek, R. (2005) The influence of locked nucleic acid residues on the thermodynamic properties of 2'-O-methyl RNA/RNA heteroduplexes Nucleic Acids Res, . 33, 5082–5093 .
Borer, P.N. (1975) Optical properties of nucleic acids, absorption and circular dichroism spectra In Fasman, G.D. (Ed.). Handbook of Biochemistry and Molecular Biology: Nucleic Acids, 3rd edn Cleveland, OH CRS Press vol. 1, pp. 589–595 .
Richards, E.G. (1975) Use of tables in calculation of absorption, optical rotatory dispersion and circular dichroism of polyribonucleotides In Fasman, G.D. (Ed.). Handbook of Biochemistry and Molecular Biology: Nucleic Acids, 3rd edn Cleveland, OH CRS Press vol. 1, pp. 596–603 .
McDowell, J.A. and Turner, D.H. (1996) Investigation of the structural basis for thermodynamic stabilities of tandem GU mismatches: solution structure of (rGAGGUCUC)2 by two-dimensional NMR and simulated annealing Biochemistry, 35, 14077–14089 .
SantaLucia, J., Jr and Turner, D.H. (1998) Measuring the thermodynamics of RNA secondary structure formation Biopolymers, 44, 309–319 .
Allawi, H.T. and SantaLucia, J., Jr. (1997) Thermodynamics and NMR of internal G·T mismatches in DNA Biochemistry, 36, 10581–10594 .
Matveeva, O.V., Shabalina, S.A., Nemtsov, V.A., Tsodikov, A.D., Gesteland, R.F., Atkins, J.F. (2003) Thermodynamic calculations and statistical correlations for oligo-probes design Nucleic Acids Res, . 31, 4211–4217 .
Crooke, S.T., Lemonidis, K.M., Neilson, L., Griffey, R., Lesnik, E.A., Monia, B.P. (1995) Kinetic characteristics of Escherichia coli RNase H1: cleavage of various antisense oligonucleotide–RNA duplexes Biochem. J, . 312, 599–608 .
Kraynack, B.A. and Baker, B.F. (2006) Small interfering RNAs containing full 2'-O-methylribonucleotide-modified sense strands display Argonaute 2/eIF2C2-dependent activity RNA, 12, 163–176 .
Ding, Y. and Lawrence, C.E. (2003) A statistical sampling algorithm for RNA secondary structure prediction Nucleic Acids Res, . 31, 7280–7301 .
Rivas, E. and Eddy, S.R. (1999) A dynamic programming algorithm for RNA structure prediction including pseudokknots J. Mol. Biol, . 285, 2053–2068 .
Wuchty, S., Fontana, W., Hofacker, I.L., Schuster, P. (1999) Complete suboptimal folding of RNA and the stability of secondary structures Biopolymers, 49, 145–165 .
Freier, S.M. and Altmann, K.H. (1997) The ups and downs of nucleic acid duplex stability: structure-stability studies on chemically-modified DNA:RNA duplexes Nucleic Acids Res, . 25, 4429–4443 .
Gryaznov, S. and Chen, J.-K. (1994) Oligodeoxyribonucleotide N3'P5' phosphoramidates: synthesis and hybridization properties J. Amer. Chem. Soc, . 116, 3143–3144 .
Pieken, W.A., Olsen, D.B., Benseler, F., Aurup, H., Eckstein, F. (1991) Kinetic characterization of ribonuclease-resistant 2'-modified hammerhead ribozymes Science, 253, 314–317 .
Heidenreich, O., Benseler, F., Fahrenholz, A., Eckstein, F. (1994) High activity and stability of hammerhead ribozymes containing 2'-modified pyrimidine nucleosides and phosphorothioates J. Biol. Chem, . 269, 2131–2138 .
Holbrook, J.A., Capp, M.W., Saecker, R.M., Record, M.T., Jr. (1999) Enthalpy and heat capacity changes for formation of an oligomeric DNA duplex: interpretation in terms of coupled processes of formation and association of single-stranded helices Biochemistry, 38, 8409–8422 .
Takach, J.C., Mikulecky, P.J., Feig, A.L. (2004) Salt-dependent heat capacity changes for RNA duplex formation J. Am. Chem. Soc, . 126, 6530–6531 .
Chaires, J.B. (1997) Possible origin of differences between van't Hoff and calorimetric enthalpy estimates Biophys. Chem, . 64, 15–23 .
Petersheim, M. and Turner, D.H. (1983) Base-stacking and base-pairing contributions to helix stability: thermodynamics of double-helix formation with CCGG, CCGGp, CGGAp, ACCGGp, CCGGUp, and ACCGGUp Biochemistry, 22, 256–263 .(Elzbieta Kierzek1,2, David H. Mathews3,4)
*To whom correspondence should be addressed. Tel: +1 585 275 3207; Fax: +1 585 276 0205; Email: turner@chem.rochester.edu
ABSTRACT
Results from optical melting studies of Watson–Crick complementary heteroduplexes formed between 2'-O-methyl RNA and RNA oligonucleotides are used to determine nearest neighbor thermodynamic parameters for predicting the stabilities of such duplexes. The results are consistent with the physical model assumed by the individual nearest neighbor-hydrogen bonding model, which contains terms for helix initiation, base pair stacking and base pair composition. The sequence dependence is similar to that for Watson–Crick complementary RNA/RNA duplexes, which suggests that the sequence dependence may also be similar to that for other backbones that favor A-form RNA conformations.
INTRODUCTION
Oligonucleotides are used for many applications, ranging from diagnostics (1–4) to therapeutics (5–9) to nanotechnology (10,11). The thermodynamics of nucleic acid duplex formation facilitates rational design of sequences for the various applications (12–14). The thermodynamics of duplex formation is dependent on the backbone of the nucleic acid. For example, the sequence dependence of the thermodynamics of DNA/DNA (15,16), RNA/RNA (17,18) and DNA/RNA (19) duplexes differ. All, however, can be approximated well by nearest neighbor models when only Watson–Crick base pairs are formed. Thus, it is relatively easy to predict the thermodynamics of Watson–Crick paired duplexes from sequence (16,18–22). Here, optical melting studies are analyzed to provide nearest neighbor thermodynamic parameters for formation of 2'-O-methyl RNA/RNA duplexes that are Watson–Crick complementary. The 2'-O-methyl RNA and other 2'-O-alkyl backbones are particularly useful for hybridization to RNA because they favor A-form helical structure and are more resistant than RNA or DNA to nuclease digestion (23–26).
MATERIALS AND METHODS
Experimental
Synthesis and purification of oligonucleotides was done as previously described (27). The buffer for melting experiments was 100 mM NaCl, 20 mM sodium cacodylate and 0.5 mM Na2EDTA, pH 7.0. Oligonucleotide single strand concentrations were determined from absorbances >80°C with extinction coefficients approximated by a nearest neighbor model (28,29). The sequence dependence of extinction coefficients for 2'-O-methyl and RNA strands was assumed to be identical. Melting curves were measured at 260 nm with a heating rate of 1°C/min from 0 to 90°C on a Beckman DU640 spectrophotometer with a water cooled Peltier thermoprogrammer. Melting curves were analyzed and thermodynamic parameters were calculated on the basis of a two-state model with the program MeltWin 3.5 (30). With one exception, agreement within 15% was found for thermodynamic parameters calculated from averaging parameters derived from the shapes of melting curves and from the following equation (20):
(1)
This agreement is consistent with the two-state model.
Parameter fitting
The measured thermodynamic parameters were fit to the individual nearest neighbor-hydrogen bonding (INN-HB) model (18) by multiple linear regression with the program Analyse-It v.1.71 (Analyse-It Software, Ltd, Leeds, England; www.analyse-it.com), which expands Microsoft Excel. Only duplexes that melted in a two-state manner were included in the fit. Measured parameters from versus ln(CT/4) plots were used as the data for the calculations. Error limits reported for the experimental data reflect the scatter in 's when fit to Equation 1. Systematic errors are typically larger, however, and difficult to estimate (18). For example, the melting is not truly two-state because the stacking in the single strand conformations is dependent on temperature and sequence. Therefore, all duplexes included in the fit were given equal weight.
RESULTS
Table 1 lists measured thermodynamic parameters for several 2'-O-methyl RNA/RNA duplexes. Only the duplex m(5'-CGAAGUGAA)/r(3'-GCUUCACUU) does not melt in an apparent two-state manner as revealed by a >15% difference between the H°s derived from averaging fits to the shapes of melting curves and from the versus ln (CT/4) plot. With the exception of m(5'-CGAAGUGAA)/r(3'-GCUUCACUU), results in Table 1 were combined with previously reported results (27) listed in Table 2 and fit to the INN-HB nearest neighbor model (18) to give the nearest neighbor thermodynamic parameters listed in Table 3.
Table 1 Thermodynamic parameters of heteroduplex formation between 2'-O-methyl RNA and oligoribonucleotides in 0.1 M NaCl, pH 7a
Table 2 Measured and predicted (in parenthesis) G°37, H°, S° and the percentage difference between measured and predicted values for Watson–Crick complementary 2'-O-methyl RNA/RNA duplexes in 0.1 M NaCl, pH 7
Table 3 Thermodynamic parameters for INN-HB nearest neighbor model applied to 2'-O-methyl RNA/RNA heteroduplexes in 0.1 M NaCl, pH 7
On a percentage basis, the errors in nearest neighbor parameters for G°37 are much smaller than for H° and S°. This is expected from the high correlation of errors in H° and S°, typically with R2 > 0.99 (18,31,32). While the individual errors in nearest neighbor parameters for H° and S° are large, the percentage errors in predicting H° and S° for duplex formation are smaller. This is because the values of H° and S° are given by the sums of the nearest neighbors, but the errors propagate as the square root of the sum of the squares of the errors. Table 2 lists the predicted values of G°37, H° and S° and their percentage differences from measured values for the oligonucleotides studied. The range of percentage differences is 0–15, 0.04–23 and 0.1–26% for G°37, H° and S°, respectively, while the average differences are, respectively, 2.4, 6.7 and 7.7%. The worst percentage prediction of G°37 differs from the measured value by 0.73 kcal/mol, which translates to a difference of 3-fold in measured and predicted association constants. The thermodynamics for the duplex m(5'-CGAAGUGAA)/r(3'-GCUUCACUU) (Table 1) are also predicted reasonably well even though it does not melt in a two-state manner and was not used in fitting the nearest neighbor parameters. The predicted values for G°37, H° and S° are –9.87 kcal/mol, –78.3 kcal/mol and –220.6 eu while the measured values are –10.56 kcal/mol, –96.6 kcal/mol and –277.6 eu, respectively.
DISCUSSION
Thermodynamic parameters for nucleic acid duplexes are useful for designing sequences for many applications, including diagnostics, therapeutics and nanotechnology (12–14,33). The 2'-O-methyl backbone and other 2'-O-alkyl backbones are particularly useful for binding to RNA because they favor A-form helixes and are chemically stable relative to DNA and RNA backbones. Thus for example, 2'-O-alkyl backbones have been used to flank a ‘gap’ DNA sequence in order to decrease nuclease digestion of the oligonucleotide while providing a long enough pairing between DNA and RNA to induce RNase H to cleave an RNA target (34). Reduction of gene expression by an RNA interference mechanism has been demonstrated with siRNA duplexes having completely 2'-O-methyl modified sense strands (35). The 2'-O- methyl modification is also used in aptamers (7), including the commercially successful therapeutic, Macugen (8).
There are several ways to analyze the data in Table 2. In principle, it is possible to fit the data to 20 parameters each for G°37, H° and S° (21,22). We chose, however, to fit the data to the 18 parameters of the INN-HB model, which ascribes separate parameters to the 16 different nearest neighbor stacks, to the average difference between a CG and UA pair, and to duplex initiation (18). This simplifies the 20 parameter model by assuming that the parameters for initiation are independent of the nature of the terminal base pairs and that terminal base pairs are equivalent to internal base pairs. With the 18 parameter fit, the highest P-values for G°37 were 0.0007 and 0.0005 for the terminal AU and m(5'-AA)/r(3'-UU) parameters, respectively. All other P-values were <0.0001. For H°, however, P-values were 0.29, 0.15 and 0.12 for the initiation, m(5'-GA)/r(3'-CU), and m(5'-CA)/r(3'-GU) parameters, respectively. After correction for salt dependence, the parameters in Table 3 can be coupled with the existing parameters for RNA loops and incorporated into dynamic programming algorithms for predicting and designing secondary structures (14,17,36–38) and can be directly compared with the 12 INN-HB parameters for RNA (18).
A comparison of the G°37 parameters for 2'-O-methyl RNA/RNA duplexes at 0.1 M NaCl in Table 3 with G°37 parameters determined for RNA/RNA duplexes at 1 M NaCl (18) shows that on average the RNA/RNA parameters for G°37 of base pair stacks are more favorable by a value of 0.26 kcal/mol (Table 3). This is likely because of higher salt concentration. The sequence dependence of stability is similar for 2'-O-methyl RNA/RNA and RNA/RNA duplexes. For example, base pair stacks with two AU pairs are less stable than base pair stacks with one AU and one GC pair; the 5'-CG/3'-GC nearest neighbor has stability similar to nearest neighbors with one AU and one GC pair, but other stacks with two GC pairs are more stable. The parameters for terminal AU pairs are 0.30 and 0.45 kcal/mol for 2'-O-methyl RNA/RNA and RNA/RNA duplexes, respectively. In the INN-HB model, this parameter is used to account for the fact that two sequences can have the same nearest neighbor base pair stacks but differ by one in the number of AU and GC pairs (18). Thus it accounts for half of the difference in hydrogen bonding free energy between AU and GC pairs. The magnitude and sign of the 0.30 kcal/mol is consistent with this physical interpretation of the model. The similarity of the sequence dependence of parameters suggests that the programs for predicting RNA secondary structure (17,36–38) can approximate 2'-O-methyl modified nucleotides as unmodified RNA, at least in base paired regions. This is because the uncertainties in the sequence dependence of loop stabilities is probably larger than the difference in sequence dependence of RNA/RNA and 2'-O-methyl RNA/RNA base pairs. Moreover, natural RNA typically has relatively few 2'-O-methyl substitutions.
The similarity of the free energy parameters for 2'-O-methyl RNA/RNA and RNA/RNA duplexes also suggests that either set may be useful for predicting the sequence dependence of stabilities of base paired duplexes with other backbones that favor A-form RNA conformations. This would include other 2'-O-alkyl, 2'-fluoro and N3'–P5'phosphoramidate backbones (24,39,40) and chimeras that include a mixture of such backbones. For example, hammerhead ribozymes and aptamers have been designed with a mixture of RNA and 2'-fluoro backbones in order to increase resistance to nuclease digestion (8,41,42). The predictions would be rough approximations, but sufficient for many applications. As expected, 2'-O-methyl RNA/RNA duplexes appear more stable than DNA/RNA duplexes. The nearest neighbour stacking parameters for 2'-O-methyl RNA/RNA in 0.1 M NaCl average 0.4 kcal/mol more stable at 37°C than those for DNA/RNA duplexes in 1 M NaCl (19).
The average error limits for the 2'-O-methyl RNA/RNA nearest neighbor parameters for G°37 and H° are 0.17 and 3.7 kcal/mol, respectively (Table 3). These are roughly 2- and 5-fold larger than the average error limits for RNA/RNA (18) and DNA/DNA (16) parameters, respectively, where only 12 parameters can be fit and the number of duplexes is larger. Direct comparison with 17 and 18 parameter fits to data for DNA/RNA duplexes is not possible because error limits for individual parameters were not reported (19,22).
The error limits for the nearest neighbor parameters for H° and S° for 2'-O-methyl RNA/RNA are large. There are many potential reasons why the nearest neighbor model does not fit the experimental H° and S° values particularly well. The two-state model for fitting the melting curve data is an approximation (43–46). The high temperature state for a duplex transition is the single strand, but stacking in these single strands is temperature dependent, thereby producing a temperature dependent H° that is not included in the model. While the agreement between H°s determined by fitting melting curves and by plotting versus ln(CT/4) is consistent with two-state behavior over the temperature range of the transition, it is not sufficient to insure two-state behavior even over this range. In addition, the INN-HB nearest neighbor model is a simple approximation. The thermodynamics may depend on more than the stacking of nearest neighbors and number of hydrogen bonds. In general, parameters obtained from optical melting data are best near the melting temperatures of the oligonucleotides. The range of melting temperatures at 0.1 mM total strand concentration for the duplexes used to determine the nearest neighbor parameters range from 28 to 55°C with an average of 44°C. Thus the predicted G°s should be reasonable from 25 to 65°C.
ACKNOWLEDGEMENTS
This work was supported by Polish State Committee for Scientific Research (KBN) grant 2 PO4A03729 to R.K., NIH grant 1RO3 TW1068 to R.K. and D.H.T., and NIH grant GM22939 to D.H.T. A.C. is a recipient of a fellowship from the President of the Polish Academy of Sciences. Funding to pay the Open Access publication charges for this article was provided by NIH.
REFERENCES
Bodrossy, L. and Sessitsch, A. (2004) Oligonucleotide microarrays in microbial diagnostics Curr. Opin. Microbiol, . 7, 245–254 .
Heller, R.A., Schena, M., Chai, A., Shalon, D., Bedilion, T., Gilmore, J., Woolley, D.E., Davis, R.W. (1997) Discovery and analysis of inflammatory disease-related genes using cDNA microarrays Proc. Natl Acad. Sci. USA, 94, 2150–2155 .
Li, H.X. and Rothberg, L.J. (2004) Label-free colorimetric detection of specific sequences in genomic DNA amplified by the polymerase chain reaction J. Am. Chem. Soc, . 126, 10958–10961 .
Long, W.H., Xiao, H.S., Gu, X.M., Zhang, Q.H., Yang, H.J., Zhao, G.P., Liu, J.H. (2004) A universal microarray for detection of SARS coronavirus J. Virol. Methods, 121, 57–63 .
Zamecnik, P.C. and Stephenson, M.L. (1978) Inhibition of Rous-sarcoma virus-replication and cell transformation by a specific oligodeoxynucleotide Proc. Natl Acad. Sci. USA, 75, 280–284 .
Milligan, J.F., Matteucci, M.D., Martin, J.C. (1993) Current concepts in antisense drug design J. Med. Chem, . 36, 1923–1937 .
Burmeister, P.E., Lewis, S.D., Silva, R.F., Preiss, J.R., Horwitz, L.R., Pendergrast, P.S., McCauley, T.G., Kurz, J.C., Epstein, D.M., Wilson, C., et al. (2005) Direct in vitro selection of a 2'-O-methyl aptamer to VEGF Chem. Biol, . 12, 25–33 .
Ruckman, J., Green, L.S., Beeson, J., Waugh, S., Gillette, W.L., Henninger, D.D., Claesson-Welsh, L., Janji, N. (1998) 2'-fluoropyrimidine RNA-based aptamers to the 165-amino acid form of vascular endothelial growth factor (VEGF165) J. Biol. Chem, . 273, 20556–20567 .
Goodchild, J. (2004) Oligonucleotide therapeutics: 25 years agrowing Curr. Opin. Mol. Ther, . 6, 120–128 .
Chworos, A., Severcan, I., Koyfman, A.Y., Weinkam, P., Oroudjev, E., Hansma, H.G., Jaeger, L. (2004) Building programmable jigsaw puzzles with RNA Science, 306, 2068–2072 .
Seeman, N.C. (2005) From genes to machines: DNA nanomechanical devices Trends Biochem. Sci, . 30, 119–125 .
Matveeva, O.V., Mathews, D.H., Tsodikov, A.D., Shabalina, S.A., Gesteland, R.F., Atkins, J.F., Freier, S.M. (2003) Thermodynamic criteria for high hit rate antisense oligonucleotide design Nucleic Acids Res, . 31, 4989–4994 .
Schwarz, D.S., Hutvagner, G., Du, T., Xu, Z.S., Aronin, N., Zamore, P.D. (2003) Asymmetry in the assembly of the RNAi enzyme complex Cell, 115, 199–208 .
Dirks, R.M., Lin, M., Winfree, E., Pierce, N.A. (2004) Paradigms for computational nucleic acid design Nucleic Acids Res, . 32, 1392–1403 .
SantaLucia, J. and Hicks, D. (2004) The thermodynamics of DNA structural motifs Annu. Rev. Biophys. Biomolec. Struct, . 33, 415–440 .
SantaLucia, J., Jr. (1998) A unified view of polymer, dumbbell, and oligonucleotide DNA nearest-neighbor thermodynamics Proc. Natl Acad. Sci. USA, 95, 1460–1465 .
Mathews, D.H., Disney, M.D., Childs, J.L., Schroeder, S.J., Zuker, M., Turner, D.H. (2004) Incorporating chemical modification constraints into a dynamic programming algorithm for prediction of RNA secondary structure Proc. Natl Acad. USA, 101, 7287–7292 .
Xia, T.B., SantaLucia, J., Burkard, M.E., Kierzek, R., Schroeder, S.J., Jiao, X.Q., Cox, C., Turner, D.H. (1998) Thermodynamic parameters for an expanded nearest-neighbor model for formation of RNA duplexes with Watson–Crick base pairs Biochemistry, 37, 14719–14735 .
Sugimoto, N., Nakano, S., Katoh, M., Matsumura, A., Nakamuta, H., Ohmici, T., Yoneyama, M., Sasaki, M. (1995) Thermodynamic parameters to predict stability of RNA/DNA hybrid duplexes Biochemistry, 34, 11211–11216 .
Borer, P.N., Dengler, B., Tinoco, I., Uhlenbeck, O.C. (1974) Stability of ribonucleic-acid double-stranded helices J. Mol. Biol, . 86, 843–853 .
Gray, D.M. (1997) Derivation of nearest-neighbor properties from data on nucleic acid oligomers.1. Simple sets of independent sequences and the influence of absent nearest neighbors Biopolymers, 42, 783–793 .
Gray, D.M. (1997) Derivation of nearest-neighbor properties from data on nucleic acid oligomers.2. Thermodynamic parameters of DNA—RNA hybrids and DNA duplexes Biopolymers, 42, 795–810 .
Cramer, H. and Pfleiderer, W. (2000) Nucleotides LXIV: synthesis, hybridization and enzymatic degradation studies of 2'-O-methyl-oligoribonucleotides and 2'-O-methyl/deoxy gapmers Nucleosides Nucleotides Nucleic Acids, 19, 1765–1777 .
Lesnik, E.A. and Freier, S.M. (1998) What affects the effect of 2'-alkoxy modifications? 1. Stabilization effect of 2'-methoxy substitutions in uniformly modified DNA oligonucleotides Biochemistry, 37, 6991–6997 .
Majlessi, M., Nelson, N.C., Becker, M.M. (1998) Advantages of 2'-O-methyl oligoribonucleotide probes for detecting RNA targets Nucleic Acids Res, . 26, 2224–2229 .
Sproat, B.S., Lamond, A.I., Beijer, B., Neuner, P., Ryder, U. (1989) Highly efficient chemical synthesis of 2'-O-methyloligoribonucleotides and tetrabiotinylated derivatives—novel probes that are resistant to degradation by RNA or DNA specific nucleases Nucleic Acids Res, . 17, 3373–3386 .
Kierzek, E., Ciesielska, A., Pasternak, K., Mathews, D.H., Turner, D.H., Kierzek, R. (2005) The influence of locked nucleic acid residues on the thermodynamic properties of 2'-O-methyl RNA/RNA heteroduplexes Nucleic Acids Res, . 33, 5082–5093 .
Borer, P.N. (1975) Optical properties of nucleic acids, absorption and circular dichroism spectra In Fasman, G.D. (Ed.). Handbook of Biochemistry and Molecular Biology: Nucleic Acids, 3rd edn Cleveland, OH CRS Press vol. 1, pp. 589–595 .
Richards, E.G. (1975) Use of tables in calculation of absorption, optical rotatory dispersion and circular dichroism of polyribonucleotides In Fasman, G.D. (Ed.). Handbook of Biochemistry and Molecular Biology: Nucleic Acids, 3rd edn Cleveland, OH CRS Press vol. 1, pp. 596–603 .
McDowell, J.A. and Turner, D.H. (1996) Investigation of the structural basis for thermodynamic stabilities of tandem GU mismatches: solution structure of (rGAGGUCUC)2 by two-dimensional NMR and simulated annealing Biochemistry, 35, 14077–14089 .
SantaLucia, J., Jr and Turner, D.H. (1998) Measuring the thermodynamics of RNA secondary structure formation Biopolymers, 44, 309–319 .
Allawi, H.T. and SantaLucia, J., Jr. (1997) Thermodynamics and NMR of internal G·T mismatches in DNA Biochemistry, 36, 10581–10594 .
Matveeva, O.V., Shabalina, S.A., Nemtsov, V.A., Tsodikov, A.D., Gesteland, R.F., Atkins, J.F. (2003) Thermodynamic calculations and statistical correlations for oligo-probes design Nucleic Acids Res, . 31, 4211–4217 .
Crooke, S.T., Lemonidis, K.M., Neilson, L., Griffey, R., Lesnik, E.A., Monia, B.P. (1995) Kinetic characteristics of Escherichia coli RNase H1: cleavage of various antisense oligonucleotide–RNA duplexes Biochem. J, . 312, 599–608 .
Kraynack, B.A. and Baker, B.F. (2006) Small interfering RNAs containing full 2'-O-methylribonucleotide-modified sense strands display Argonaute 2/eIF2C2-dependent activity RNA, 12, 163–176 .
Ding, Y. and Lawrence, C.E. (2003) A statistical sampling algorithm for RNA secondary structure prediction Nucleic Acids Res, . 31, 7280–7301 .
Rivas, E. and Eddy, S.R. (1999) A dynamic programming algorithm for RNA structure prediction including pseudokknots J. Mol. Biol, . 285, 2053–2068 .
Wuchty, S., Fontana, W., Hofacker, I.L., Schuster, P. (1999) Complete suboptimal folding of RNA and the stability of secondary structures Biopolymers, 49, 145–165 .
Freier, S.M. and Altmann, K.H. (1997) The ups and downs of nucleic acid duplex stability: structure-stability studies on chemically-modified DNA:RNA duplexes Nucleic Acids Res, . 25, 4429–4443 .
Gryaznov, S. and Chen, J.-K. (1994) Oligodeoxyribonucleotide N3'P5' phosphoramidates: synthesis and hybridization properties J. Amer. Chem. Soc, . 116, 3143–3144 .
Pieken, W.A., Olsen, D.B., Benseler, F., Aurup, H., Eckstein, F. (1991) Kinetic characterization of ribonuclease-resistant 2'-modified hammerhead ribozymes Science, 253, 314–317 .
Heidenreich, O., Benseler, F., Fahrenholz, A., Eckstein, F. (1994) High activity and stability of hammerhead ribozymes containing 2'-modified pyrimidine nucleosides and phosphorothioates J. Biol. Chem, . 269, 2131–2138 .
Holbrook, J.A., Capp, M.W., Saecker, R.M., Record, M.T., Jr. (1999) Enthalpy and heat capacity changes for formation of an oligomeric DNA duplex: interpretation in terms of coupled processes of formation and association of single-stranded helices Biochemistry, 38, 8409–8422 .
Takach, J.C., Mikulecky, P.J., Feig, A.L. (2004) Salt-dependent heat capacity changes for RNA duplex formation J. Am. Chem. Soc, . 126, 6530–6531 .
Chaires, J.B. (1997) Possible origin of differences between van't Hoff and calorimetric enthalpy estimates Biophys. Chem, . 64, 15–23 .
Petersheim, M. and Turner, D.H. (1983) Base-stacking and base-pairing contributions to helix stability: thermodynamics of double-helix formation with CCGG, CCGGp, CGGAp, ACCGGp, CCGGUp, and ACCGGUp Biochemistry, 22, 256–263 .(Elzbieta Kierzek1,2, David H. Mathews3,4)