- Citado por SciELO
versión On-line ISSN 0717-9707
J. Chil. Chem. Soc. v.50 n.1 Concepción mar. 2005
J. Chil. Chem. Soc., 50, N 1 (2005)
" INTERACTION ENERGIES IN NON WATSON-CRICK PAIRS: AN AB INITIO STUDY OF G·U AND U·U PAIRS "
SANDRA T. MADARIAGA1, J.GUILLERMO CONTRERAS2 and C.GLORIA SEGUEL3
1)Centro de Docencia Superior en Ciencias Basicas, Universidad Austral de Chile, Casilla 1327, Puerto Montt, Chile.
Ab initio calculations at the MP2/6-31G** level have been carried out on the non-Watson-Crick nucleic acids pairs G·U and U·U to obtain the interaction energies and to see whether the derived values are comparable or not with the canonical G-C, A-T and A-U pairs. Optimized geometries of the pairs show that the structural parameters of the isolated bases differ very little on pairing. The guanine -NH2 group does not participate in the hydrogen bonding formation and possesses a pyramidal structure; its intrinsic nonplanarity plays an important role in the out-of-plane intermolecular interactions. Thus, the G·U pair projects three hydrogen bonding acceptor sites, namely, N7(G), O6(G) and O4(U) to the RNA major groove. The interaction energy (DHoint) calculated for the G·U pair (-13.6 kcal/mol) is comparable to that determined for A-T (-13.0 kcal/mol), but considerable smaller than the experimental value reported for G-C (-21.0 kcal/mol). The U·U pair follows the trend that pairing between pyrimidines bases should have lower interacting energies than purine-pyrimidine pairs.
Keywords: Ab initio; interaction energies; G·U , U·U pairs.
Investigations in molecular biology1-4 show that the three dimensional structure of the different RNAs is formed by non-canonical base pairs (non-Watson-Crick or NWC) in addition to the classical Watson-Crick ( WC pairs) A-U and G-C. In fact the G·U pair is largely one of the most important of such as elements and participates in RNA-RNA and RNA-protein interactions. In the later case5-8, the G·U pair produces a distortion in the RNA regular helix structure and provides the electron donor and acceptor groups for intermolecular contact with aminoacids. In some cases , such as in r-RNA, the G·U pair is replaced by U·U producing changes in the overall geometry of the double helix9,10. Although, no detailed information about the ribosomal RNAs functions, is available, it is known that they act as catalysts. Thus, it is believed that the G·U participates in the ribosomal catalytic core formation11. In DNA, the NWC pairs are potential points of carcinogenesis. However, in vitro studies reveal the presence of non-conventional associations between the nitrogenated bases12. It is known that a number of NWC interactions involving different edges of the basis occur inside the different RNAs and that the position and orientations of the electron donor and acceptor groups, so important in the hydrogen bonding formation, can be inferred from the x-ray diffraction or 1H-NMR data13-15. The purine and pyrimidine bases of nucleic acids show three edges to form hydrogen bonding between them or other molecules: a) the Watson-Crick edge, b) the Hoogsteen edge (for purines) or CH edge (for pyrimidines) and c) the sugar edge. Interactions through any of these edges produce non-canonical pairs in the RNAs. In fact, Leontis et al16 have proposed twelve «families» of non-canonical pairs, though the main factors governing the different NWC interactions are still emerging. It is believed that the nucleosides generated from these types of interaction will be the preferred recognition sites of certain molecules binding to the RNAs. Thus, the Watson-Crick (WC) pairs G-C and A-U, are considered as the basic units of the tridimensional structure of these biomolecules whereas the NWC pairs lead to a highly structured RNA and their presence is of great importance in the molecular recognition processes associated to the RNA functions. These pairs make possible the RNA-RNA, RNA-protein interactions and are the sites for drugs recognition and ions5,13. The fidelity of the protein syntheses depends on the specific amino-acylation, i.e., on the covalent bonding formation between and aminoacid and a base of the t-RNA catalyzed by the aminoacyl-t-RNA- synthethase. The action of this enzyme depends of the recognition of some nucleotides and structural characteristics in the t-RNA. Thus, the G·U pair plays an important role17,18 and has been identified in a number of RNA structures6,19. In other words, the presence of the G·U pair assures the amino-acylation process. If guanine is replaced by inosine in the G·U pair, the acylation process does not occur due to the absence of the 2-amino group20 in the later. It has been suggested that the AC, GA and CC pairs show functional characteristics comparable to the G·U pair21-23. In fact, x-ray diffraction studies show that N1 in adenine is protonated and acts as a donor center in the hydrogen bonding to cytosine24,25. In this sense the AC pair is structurally similar to the G·U pair. Many RNA molecules work with proteins in a concerted way by forming permanent (ribosomes) and transient complexes such as m-RNA. Due to the larger number structural blocks, proteins are better and more specific catalysts than RNAs26. The RNA-protein complexes require the precise mutual molecular recognition. The NWC pairs participate in this process producing the interactions with the aminoacids of the polypeptide chains. Thus, the lysine residue forms hydrogen bonds to O4 of both uracils in the U·U pair. In addition to well known hydrogen bonding between the nucleic acids bases, into of some ribosomes, bifurcated bonding, i.e., two hydrogen interact to a single donor center as in the G·U and GG non-canonical pairs16. In the present work the structural, electrostatic and energetic characteristics of the G·U and U·U pairs are analyzed. Accordingly, ab initio molecular orbital calculations, both at the Hartree-Fock and Møller-Plesset levels and the 6-31G** basis set, have been carried out in an attempt to compare the present results with those found for the canonical G-C, A-T and A-U pairs.
Standard ab initio calculations, in the frame of molecular orbital theory, were performed using Gaussian 98 suite of programs27. Geometry optimization for the two non-canonical pairs were carried out at HF/6-31G** level, followed of frequency calculations in order to see whether the derived structures were local minima. Frequency calculations at HF/6-31G** level indicate that the noncanonical pairs show imaginary frequencies when the angles involving the hydrogen bonding are ca. 180. The natural pairs G-C and A-T are local minima for such as angles close to 180 28. Energy calculations at MP2/6-31G**, to include electron correlation in the frozen approximation, were carried out and corrected for vibrational ZPE. To obtain enthalpies, ZPE and thermal correction (H-Ho) were added to the energy calculated at MP2/6-31G**. Interaction energies were obtained from the E value for the pair to which the energies of the isolated purine and pyrimidine bases have been subtracted. The same procedure was applied to obtain DZPE and D(H-Ho) values. BSSE (basis set superposition error) corrections were calculated using the counterpoise method (CPM) formulated by Boys and Bernardy29. Accordingly, BSSE(XY) = E(X)X + E(Y)Y - E(X)XY - E(Y)XY, where E(X)XY is the energy of X calculated at the supermolecule basis set, E(X)X corresponds to the energy of X calculated at its own basis set. It is known that CPM works well at HF level, whereas conflictive opinions have been reported30,31 on the application to correlated methods. The application of BSSE corrections to the MP2/6-31G** level will be discussed later on. The solute-solvent interactions were modeled using the polarized continuum method (PCM)32,35. The method behaves well in reproducing experimental free energies of solvation (DGoS ) in water for a number of simple molecules36. AGoS contains both electrostatic and non-electrostatic terms (cavitation, dispersions and repulsion energies). Since structural parameters change very little on going from the gas phase to solution no large effect on DGoS is expected37,38. Accordingly, to obtain the solvation free energies the gas phase geometries were used.
RESULTS AND DISCUSSION
The G·U and U·U pairs correspond to molecular interactions between the WC edges in the former and the Hoogsteen edges in the later, respectively. The molecular recognition implies two N-HO hydrogen bonding as it is seen in the HF/6-31G** optimized structures (see figure 1). The atom numbering used in all tables is that shown in this same figure. To obtain the corresponding pairs, the most stable tautomeric forms of guanine (G) and uracil (U) were used28. Table 1 lists some relevant bond distances and angles for both the isolated and paired G and U bases.
From table 1 it can be inferred that in general the structural parameters do not change greatly upon pairing. This observation is consistent with previously reported structures
Fig. 1: HF/6-31G** optimized structures of the U·U and G·U pairs
on natural and non-natural WC pairs28,39. The non-bonded distances (table 2) between the electronegative centers (O.N) are ca 3.0 Å and are in good agreement with the values found by diffraction methods1. In the "pairing zone", the N-H.O bond angles in G·U and U·U non-canonical pairs, are in the range of 167 - 176. This range is clearly different of the one found in the WC pairs (175-178)28,40. An interesting feature of the G·U pair is the
non-planarity of the amino group of G, that corresponds to a local minimum in the optimized geometry. The planarity of this group yields an imaginary frequency and thereby the structure would be a transition state. Ab initio calculations predict that the amino group is slightly pyramidal in the isolated bases but planar in the complementary pairs formed by purines and azapurines28,41-43. The non-planarity determine the attractive intermolecular interactions of the hydrogen atoms of the -NH2 group oriented out of plane of the nucleic acids bases44. In fact, the experimental data21-23,44,45, confirm that the exocyclic amino group acts in the amino acylation of the t-RNA17,19 and plays an important role in the catalytic core formation at the ribozyme11. The main difference between canonical and no-canonical base pairs is related with the type and location of functional group into RNA. The G·U pair projects the NH2 group to the minor groove, whereas in the G-C and A-U WC pairs ,this group points towards the major groove8 and is being used in the hydrogen bonding formation with cytosine and Uracil. On the other hand, the G·U pair projects three acceptor groups towards the mayor groove, namely, N7 (G), O6(G) and O4(U). Calculations at HF/6-31G** level reveal that the atomic charges are ca. -0.53, -0.66 and -0.50, respectively. The charge distribution inside the heterocycles is relevant parameters to be considered in the study of complementary bases since the interaction energies depend on the partial charges of the atoms. In general, it has been observed that the atoms involved in the hydrogen bonding acquire small additional charges after pairing39. Accordingly, the nitrogen atom of the NH2 group in isolated G shows a charge of -0.76 changing to -0.78 in noncanonical pair. In the WC G-C pair, the -NH2 group is forming a hydrogen bonding with cytosine and possesses an atomic charge of ca. -0.8228 that is comparable with the value calculated in A-U pair (ca. 0.80), where adenine is forming a hydrogen bonding with uracil.
Table 3 shows the calculated interaction energies (DHoint. ) for the G·U, U·U and A-U pairs. These energies have been corrected for ZPE, H-Ho and BSSE. The BSSE have been calculated at both HF/6-31G** and MP2/6-31G** levels . The BSSE values derived at MP2 are ca. twice the calculated at the Hartree-Fock level consistent with previously reported findings28,39. However, the calculated interaction energies derived at HF and MP2 levels, for both non-canonical pairs differ just in 1.30 and 2.05 kcal/mol. The canonical A-U pair shows an energy interaction larger than U·U but 1.2 kcal/mol smaller than G·U. On the other hand, it is known that the SCF/BSSE corrections to the energies yield interaction energies comparable to the experimental values. Thus, for the canonical pair G=C, the calculated and experimental DHoint. are ca. -21.3 and -21.0 kcal/mol, respectively, whereas for the A-T pair, these values46 are ca. -12.0 and -13.0 kcal/mol, respectively. From the above interaction energies, it can be inferred that the pairing in U·U follows the prediction that interactions between pyrimidine bases must be smaller than a purine-pyrimidine one.
The polarized continuum model (PCM)32-35 was used to study the solvent effect on the pairing energies. PCM predicts the G·U solvation free energy (see table 4) to be ca -26.3 kcal/mol, i.e., ca. 2.0 kcal/mol larger than the DGoS found for the natural pair G-C. For U·U, DGoS is ca 7.0 kcal/mol smaller than G·U. The WC G-C pair is ca. 5.0 kcal/mol more stable in solution than A-T and U·U. The canonical A-U pair presents an aqueous solvation free energy and a dipole moment comparable with the A-T pair. The theoretical energies in solution as predicted by PCM are consistent with the experimental data. In fact, a value of -1.7 kcal/mol found spectroscopically47 , compares well with our value of ca. -1.44 kcal/mol. The energies per hydrogen bonding calculated by PCM differ from the values reported by Stofer et al. These energies are of ca. 2.0 kcal/mol in G-C, whereas values of ca. 0.58, 0.72 and 1.20 kcal/mol have been found for A-T, U·U and G·U, respectively. The energy for hydrogen bonding is ca. 0.57 kcal/mol in A-U and represents a difference of just 0.01 kcal/mol with respect to that found for the A-T pair. Ab-initio calculations clearly show the pairs studied in the present work possess similar hydrogen bonding energies and that the structural features are in agreement with spectroscopic studies.
1.- The ab-initio calculation at HF/6-31G** level show that the struc tural parameters do not change upon pairing and that NH2 group is not planar in the G·U pair. Accordingly, in the RNA, the G·U pair participates in the amino-acylacion process.
This work was supported by operating grants (No S-200168) from the Universidad Austral de Chile and (No 203.021.019-1.0) from the Universidad de Concepcion (Chile).
1.- S.R. Price, P.R. Evans, K. Nagai, Nature 394 (1998) 645-650. [ Links ]
2.- M. Wu, I. Tinoco Jr., Biochemistry 95 (1998) 11555-11560. [ Links ]
3.- G.L. Conn, D.E. Draper, E.E Lattman, A.G Gittis, Science 284 (1999) 1171-1174. [ Links ]
4.- N.B. Leontis, J. Stombaugh, E. Westhof, Nucleic Acids Res. 30 (2002) 3497-3531. [ Links ]
5.- T. Hermann, E. Westhof, Chemistry and Biology 6 (1999) R335- R343. [ Links ]
6.- S.A. Strobel, L. Ortoleva-Donnelly, S.P. Ryder, J.H. Cate, E. Mancoeur, Natl. Struct. Biol. 5 (1998) 60-66. [ Links ]
7.- M. Szyma, M.Z. Barciszewska, V.A. Erdmann, J. Barciszewski, Mol. Biol. and Evol., 17 (2000) 1194-1198. [ Links ]
8.- G. Varani, W.H. McClain, EMBO reports 1 (2000) 18-23. [ Links ]
9.- A. Dallas, P.B. Moore, Structure 5 (1997) 1639-1653. [ Links ]
10.- K.J, Baeyens, H.L, De Bondt, S.R, Holbrok, Natl. Struct. Biol. 2 (1995) 56-62. [ Links ]
11.- S.A. Strobel, T.R. Cech, 1995, Science 267 (1995) 675-679. [ Links ]
12.- G.J. Quigley, A.R. Rich, Science 194 (1975) 796-806. [ Links ]
13.- J. Wohnert, A.J. Dingley, M. Stoldt, M. Gorlach, S. Grzesiek, L. Brown, Nucleic cid Res. 27 (1999) 3104-3110. [ Links ]
14.- B. Masquida, E. Westhof, RNA 6 (2000) 9-15. [ Links ]
15.- X. Che, J.A. McDowell, R. Kierzek, T.R. Krug, D.H. Turner, Biochemistry 39 (2000) 8970-8982. [ Links ]
16.- N.B. Leontis, E. Westhof, RNA 7 (2001) 499-512. [ Links ]
17.- K. Musier-Forsyth, N. Usaman, S. Scaringe, J. Doudna, R. Green, P. Schimmel, Science 253 (1991) 784-786. [ Links ]
18.- M. Frugier, P. Schimmel, Proc. Natl. Acad. Sci. USA., 94 (1997) 11291-11294. [ Links ]
19.- D. Gautheret, D. Konings, R. R. Gutell, RNA 1 (1995) 807-814. [ Links ]
20.- P.J. Beuring, F. Yang, P. Schimmel, K. Musier-Forsyth, Proc. Natl. Acad. Sci. USA. 94 (1997) 10150-10154. [ Links ]
21.- Kung-Yao Chang, G. Varani, S. Bhattacharya, H. Choi, W.H. McClain, Proc. Natl. Acad. Sci. USA. 96 (1999) 11764-11769. [ Links ]
22.- W.H. McClain, K Gabriel, S. Bhattacharya, Y.Y. Jou, J. Schneider, J. Mol. Biol. 286 (1999) 1025-1032. [ Links ]
23.- K. Gabriel. J. Schneider, W.H. McClain, Science 271 (1996) 195-197. [ Links ]
24.- S.B. Jang, L.W. Hung, Y.I. Chi, E.L. Holbrook, R.J. Carter, S.R. Holbrook, Biochemistry 37 (1998) 1726-1731. [ Links ]
25.- U. Nagaswamy, N. Voss, Z. Zhang, G.E. Fox, Nucleic Acids Res. 28 (2000) 375-376. [ Links ]
26.- J. A. Piccirilli, T. Krauch, S. E. Moroney, S.A. Benner, Nature 343 (1990) 33-37. [ Links ]
27.- M. J. Frisch, G.V. Trucks, H.B. Schlegel, M.A. Scuseria, M.A. Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery Jr., et-al, R.E. Stratman, J.C. Burant, D. Dapproch, J.M. Millan, A.D. Daniels, K.N. Kudin, M.C. Stain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Menucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G.A. Peterson, P.Y. Ayala, Q. Cui, K. Morokuma, D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman, J. Cioslowski, J.B. Ortiz, A.G. Baboul, B.B. Stefanov, G. Lui, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L. Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A. Nanayakkara, C. Gonzalez, M. Chalacombe, P.M.W. Gill, B. Johson, W. Cheng, M.W. Wong, J.L. Andres, M. Head-Gordon, E.S. Repogle, J.A. Pople, Gaussian 98, revision A7, Gaussian Inc., Pittsburg, PA., USA, 1998. [ Links ]
28.- J. G. Contreras, S.T. Madariaga, Bioorg. Chem. 31 (2003) 367- 377. [ Links ]
29.- S. F. Boys, F. Bernardi, Mol. Phys. 19 (1970) 553-557. [ Links ]
30.- P. Hobza, R. Zahradnik, Chem. Rev. 88 (1988) 871-897. [ Links ]
31.- D. W. Schewenke, D. G. Truhlar, J. Chem. Phys. 82 (1985) 2418- 2426. [ Links ]
32.- S. Miertus, J. Tomasi, J. Chem. Phys. 65 (1982) 239-246. [ Links ]
33.- M. T. Cossi, V. Barone, R. Cammi, J. Tomasi, J. Chem. Phys. Lett. 255 (1996) 327-333. [ Links ]
34.- M. T. Cances, V. Menuccini, J. Tomasi, J. Chem. Phys. 107 (1997) 3031- 3037. [ Links ]
35.- V. Barone, M. Cossi, J. Tomasi, J. Chem. Phys. 107 (1997) 3219- 3227. [ Links ]
36.- J. Tomasi, M. Persico, Chem. Rev. 94 (1994) 2027-2033. [ Links ]
37.- J. G. Contreras, S. T. Madariaga, J. Mol. Struct. (THEOCHEM) 572 (2001) 235-242. [ Links ]
38.- J. G. Contreras, S.T. Madariaga, J. Phys. Org. Chem. 13 (2000) 187-191. [ Links ]
39.- S.P Jiang, G. Raghunathan, K.L. Ting, J.C. Xuan, R.L. Jernigan, J. Biomol. Struct.& Dyn. 12 (1994) 12:367-381. [ Links ]
40.- S.T. Madariga, J.G. Contreras, J. Chilean Chem. Soc. 48 (2003) 129-133. [ Links ]
41.- P. Hobza, J. Sponer, Chem. Rev. 99 (1999) 3247-3276. [ Links ]
42.- D. Barsky, E. Kool, M. Colvin, J. Mol. Struct. & Dyn. 16 (1999) 1119-1133. [ Links ]
43.- J. Sponer, I. Berger, N. Spackova, J. Leszczynski, P. Hobza. J. Biomol. Struct. & Dyn. ,17 (2000) 383-407. [ Links ]
44.- J. Sponer, J. Leszcznski, P. Hobza. J. Biomol. Struct. & Dyn. 14 (1996) 117-135. [ Links ]
45.- P.J. Beuring, F. Yang, P. Schimmel, K. Musier-Forsyth, Proc. Natl. Acad. Sci. USA. 94 (1997) 10150-10154. [ Links ]
46.- I. K. Yanson, A. R. Teplitsky, L. F. Sukhodub, Biopolymers 18 (1979) 1149-1153. [ Links ]
47.- E. Stofer, C. Chipot, R. Lavery, J. Am. Chem. Soc. 121 (1999) 9503-9508. [ Links ]
Todo el contenido de esta revista, excepto dónde está identificado, está bajo una Licencia Creative Commons