8680
2007, 111, 8680-8683 Published on Web 07/07/2007
Close Contacts between Carbonyl Oxygen Atoms and Aromatic Centers in Protein Structures: π‚‚‚π or Lone-Pair‚‚‚π Interactions? Alok Jain,† Chandra Shekhar Purohit,‡ Sandeep Verma,‡ and Ramasubbu Sankararamakrishnan*,† Department of Biological Sciences and Bioengineering, and Department of Chemistry, Indian Institute of TechnologysKanpur, Kanpur-208016 (UP), India ReceiVed: April 8, 2007; In Final Form: June 4, 2007
Lone-pair‚‚‚π and, more recently, π‚‚‚π interactions have been studied in small molecule crystal structures, and they are the focus of attention in some biomolecules. In this study, we have systematically analyzed 500 high-resolution protein structures (resolution e1.8 Å) and identified 286 examples in which carbonyl oxygen atoms approach the aromatic centers within a distance of 3.5 Å. Contacts involving backbone carbonyl oxygens are frequently observed in helices and, to some extent, in strands. Geometrical characterization indicates that these contacts have geometry in between that of an ideal π‚‚‚π and a lone-pair‚‚‚π interaction. Quantum mechanical calculations using 6-311++G** basis sets reveal that these contacts give rise to energetically favorable interactions and, along with MD simulations, indicate that such interactions could stabilize secondary structures.
Weak dispersive interactions such as nonconventional hydrogen bonding,1-3 cation‚‚‚π interactions,4,5 anion‚‚‚π interactions,6 and π‚‚‚π interaction7 are widely regarded as stabilizing factors for a variety of macromolecular structures, supramolecular assemblies, and molecular recognition.8,9 Cation‚‚‚π interactions are known to play a crucial role in ligand recognition and catalysis,10 stabilizing protein-protein interfaces,11 and ion selectivity of potassium channels.12 Anion‚‚‚π interactions take part in molecular recognition, namely, halide receptor specificity.13 Lone-pair‚‚‚π (lp‚‚‚π) interactions, an intermediate class between the above two, by electronic considerations, has been a subject of debate. Elegant theoretical investigations by Dougherty and co-workers14 have provided ample proof for the existence of such an interaction, despite being counterintuitive from an electrostatic point of view. In the H‚‚‚π interaction, the LUMO of the water and the HOMO of the aromatic ring are involved, and it is in stark contrast to the lp‚‚‚π interaction where the HOMO of the water and the LUMO of the aromatic rings interact.15 Recently, a solid-state structure analysis showed the presence of lp‚‚‚π interactions between a pentafluorophenyl group and an alcoholic oxygen on the basis of distance criteria.16 This type of interaction has also been reported between a water molecule and functionally important unstacked cytosine and adenosine bases in a high-resolution crystal structure of a ribosomal frameshifting RNA pseudoknot from a beet western yellow virus.17 A recent study identified the existence of such interactions in a number of small-molecule crystal structures from the Cambridge Structural Database (CSD) and also in DNA and RNA struc* To whom correspondence should be addressed. E-mail: rsankar@ iitk.ac.in. Telephone: (+91)-512-2594014. Fax: (+91)-512-2594010. † Department of Biological Sciences and Bioengineering. ‡ Department of Chemistry.
10.1021/jp072742l CCC: $37.00
tures.18 The same study also concluded that lp‚‚‚π interactions can occur only rarely in protein structures. However, a systematic analysis of close contacts between carbonyl oxygen atoms and an aromatic centroid has not been carried out in protein structures. In this paper, we have investigated the occurrence of such contacts in 500 high-resolution protein structures (resolution e1.8 Å).19 The following geometric criteria were used to identify carbonyl oxygen-aromatic center contacts in proteins. At first, all carbonyl oxygen atoms (O) lying within a distance of 3.5 Å from the center of the aromatic residues (histidine, tyrosine, phenylalanine, and both rings of tryptophan) were retrieved. If the occupancy value of atoms was not one, then the residues were discarded. Further filtering was achieved by excluding those aromatic residues that participate in hydrogen bond interactions through their imidazole or indole nitrogen donors with the O atom acting as the acceptor. This was done to ascertain that the proximity of oxygen to the aromatic residue was not due to any constraints. On the basis of these criteria, we have identified 286 examples in which carbonyl oxygen atoms were in close proximity to the center of one of the four aromatic residues. This set of contacts was further subdivided into two categories, contacts involving main-chain carbonyl oxygen (Figure 1a) and those with side-chain carbonyl oxygen (Figure 1b). Among the four aromatic residues, aromatic centers of His and, to some extent, Trp seem to be more frequently involved in close contacts with oxygens relative to their overall occurrence (Figure 1c). A similar observation using ab initio calculations has recently been made on protonated imidazole species.20 The distribution of angles involving the O atom, the ring centroid (RC) and one of the aromatic carbons (AC) (Figure 1d), exhibits a preference for 70-110°. Analysis of small© 2007 American Chemical Society
Letters
J. Phys. Chem. B, Vol. 111, No. 30, 2007 8681
Figure 2. Potential energy plot for the benzene-formaldehyde complex.
Figure 1. Different types of oxygen-aromatic center (RC) contacts, (a) RC‚‚‚backbone O (249 examples), (b) RC‚‚‚side chain O (37 pairs), and (c) histogram representing the distribution of individual aromatic amino acids in the whole aromatic data set (blue) and those involved in oxygen-aromatic contacts (red). (d) Scatter plot of an oxygenaromatic centroid contact versus the O-RC-AC angle (see text). (e) Histogram showing the distribution of dihedral angles ω between the planes formed by the X2CdO group and the aromatic ring. (f) Residue separation between the aromatic residues and the residues that provide contacting carbonyl oxygens.
molecule structures in the CSD database indicates a similar trend in anion‚‚‚π interactions observed in a number of chemical compounds.6 One could argue that the CdO‚‚‚aromatic centroid contacts are due to the π‚‚‚π interactions involving the π electrons of the CdO double bond/peptide bond and the aromatic ring. One of the ways to resolve this issue is to calculate the dihedral angle (ω) between the planes defined by the X2CdO and the aromatic ring, where X denotes any atom.18 When ω ) 0°, the carbonyl oxygen will be stacked on the aromatic plane and is considered ideal for π‚‚‚π interactions. At ω ) 90°, the CdO group will have a head-on orientation toward the ring centroid that will indicate the existence of lp‚‚‚π interactions. The distribution of ω in our data set is between 25 and 65° (Figure 1e), indicating that the carbonyl group approaches the aromatic ring at an angle. Thus the oxygen-aromatic centroid contacts could be due to weak π‚‚‚π interactions. However, this analysis shows that lp‚‚‚π interactions involving the lone pairs of oxygen atom and the π electrons of the aromatic ring cannot be completely ruled out. Ab initio quantum mechanical (QM) calculations (Gaussian 03 suite of programs)21 on model compound benzeneformaldehyde were carried out to investigate the favorable nature of such contacts. It is known that the size of the basis set affects the sensitivity of the result.22 Hence, we have used a large basis set for all of our calculations, with the aim of getting relatively accurate results. Initially, monomers were individually optimized at the MP2 level of theory using 6-311++G** basis sets. Subsequently, a single point step scan with the same level of theory was carried out using initial geometries where the
formaldehyde plane makes a dihedral 0, 35, 45, 55, and 90° angle with the benzene plane. The distance between the benzene center and O of formaldehyde was varied in steps of 0.05 Å for 100 different points, resulting in a one-dimensional potential energy diagram (Figure 2). BSSE-corrected interaction energies at the 0, 35, 45, 55, and 90° orientations were -2.88, -1.42, -1.19, -1.05, and -0.93 kcal/mol, respectively (Table S1 of Supporting Information), indicating that CdO‚‚‚aromatic center contacts occurring at an angle of ∼45° are indeed favorable (see Supporting Information). However, the interaction is the most favorable when there is π‚‚‚π interaction (ω ) 0°). Even when ω ) 90°, our calculations indicate that there could be favorable lp‚‚‚π interaction. In the majority of examples in our data set, ω values varied between 35 and 55°, in between that of an ideal π‚‚‚π and lp‚‚‚π interaction. Bader’s theory of “atoms in molecules” (AIM) provides an unambiguous definition of chemical bonding23,24 and gives some useful information regarding the strength of the noncovalent interactions in the complexes such as the one currently under investigation. AIM theory has been successfully used to characterize anion‚‚‚π interactions in several complexes.6,25 In the present study, we have determined the topological analysis of the charge density and properties of critical points (CPs) for a benzene-formaldehyde complex in three different geometrical arrangements in which the dihedral angle, ω, between the planes formed by the formaldehyde and benzene assumed 0, 45, and 90° orientations. The calculations were performed using the AIM2000 program26 at the MP2/6-311++G** level of theory. When ω ) 0 or 45°, the exploration of CPs revealed one (3, -1), one (3, +1), and one (3, +3) CPs (Figure 3a,b). In the case of ω ) 0°, the bond CP connects the carbonyl carbon of HCHO with one of the aromatic carbons, indicating the existence of a π‚‚‚π interaction. When ω ) 45°, the CPs shows a different pattern of connection. The bond CP connects the carbonyl oxygen of HCHO with one of the aromatic carbons, and the ring CP connects the oxygen of formaldehyde with the aromatic carbon that is diametrically opposite to the one connected by the bond CP. The pattern of interaction indicates that the lone pairs of oxygen seem to have an influence in the interaction between formaldehyde and benzene. When ω ) 90°, the exploration of CPs reveals the presence of two (3, -1), two (3, +1), and one (3, +3) CPs (Figure 3c). The two bond CPs connect the carbonyl oxygen of formaldehyde with two aromatic carbons that are located in the opposite sides of the aromatic plane. One ring CP connects the oxygen with the midpoint of two aromatic carbons located in between the two aromatic carbons that are connected by the bond CPs. It should be mentioned that ω ) 90° is an ideal orientation for the lp‚‚‚π interaction. The Laplacian of the (3, -1) CPs is positive in all three cases (Table S2 of Supporting Information), indicating a depletion of the electron density, as is common in closed-shell interactions. BSSE-corrected interaction energies and the pattern
8682 J. Phys. Chem. B, Vol. 111, No. 30, 2007
Letters
Figure 5. Side-chain conformations of initial structures of the peptides considered for simulation, (a) System-I, χ1 ) -55° and (b) System-II, χ1 ) 170°; MD distance trajectories generated for a 15 ns production run between the center of aromatic ring and the backbone carbonyl oxygen for (c) System-I and (d) System-II.
Figure 3. Schematic representation of the location of the (3, -1) (red circles), (3, +1) (yellow circles), and (3, +3) CPs (green circles) originating from the interaction of a benzene aromatic ring with the CdO group of formaldehyde at different orientations, (a) ω ) 0°, (b) ω ) 45°, and (c) ω ) 90° from two points of view. Model complexes are shown in the side-view (left) and perpendicular to the benzene aromatic ring (right). The formaldehyde has been omitted for the sake of clarity in the top views on the right.
Figure 4. Representative examples of (a) i f i + 4 in the helix and (b) i f i ( 1 in the strand.
of connectivity of CPs indicate that the interaction between the carbonyl oxygen and the aromatic center observed in the protein structures will have contributions from both π‚‚‚π and lp‚‚‚π interactions. In 82 cases, backbone O of an amino acid is in close proximity to the center of an aromatic residue and is separated by four residues (Figure 1f). These contacts can be described as an i f i + 4 type, and they are overwhelmingly observed in R-helices (Figure 4a). Similar trends for other interactions involving a π electronic cloud such as cation‚‚‚π, π‚‚‚π, and X-H‚‚‚π have also been reported, and they are shown to provide extra stability for helices.2,27-29 An aromatic ring can come into contact with the backbone O of its preceding or succeeding residue. This kind of i f i ( 1 interaction is the second most favored type (Figure 1f). However, it is less frequently found in β-strands (Figure 4b).
We performed molecular dynamics (MD) simulations (using GROMACS suite)30 on a 27 residue polyleucine model helical peptide to examine the stability of oxygen-aromatic contacts. An aromatic residue, Tyr, in the middle of this helix was introduced, which contacted the backbone O one turn below (System-I; Figure 5a). This system was simulated for a period of 15 ns. By changing the side-chain dihedral angle of Tyr, the oxygen-aromatic contact was disrupted (System-II; Figure 5b), and this system was subjected to another MD simulation. In System-I, the oxygen-aromatic contact remained stable throughout the simulation (Figure 5c). Interestingly in SystemII, after 2 ns, a transition in the side-chain dihedral angle of Tyr brought the aromatic ring closer to the backbone O of the residue located one helical turn below, re-establishing the contact similar to that for System-I. For the rest of the simulation, this interaction was observed almost uninterrupted (Figure 5d), demonstrating its stable nature in force fields used in classical MD simulations. In summary, we have identified a large number of examples in protein structures in which carbonyl oxygens are in close proximity to the aromatic centers. The contacts observed predominantly in helices and strands involve the backbone CdO groups. Subsequent QM and MD studies indicate that the interactions from these contacts are favorable and stable. The geometry defining the contacts between CdO and the aromatic centroid lies in between that of an ideal π‚‚‚π and an ideal lp‚‚‚π interaction, and the interactions arising from these contacts could have contributions from both. Our calculations support the idea that the interactions between the carbonyl oxygen and the aromatic centers could play a stabilizing role in protein secondary structures. Acknowledgment. A.J. and C.S.P thank IIT-Kanpur and CSIR, India, for a research fellowship. We gratefully acknowledge Professor N. Sathyamurthy and Mr. Brijesh Mishra for their help in AIM calculations. S.V. thanks DST for a Swarnajayanti fellowship in chemical sciences. R.S. thanks IBM for the computational facility made available through its SUR grant. We thank Tarini Shankar Ghosh for his help during the course of this study.
Letters Supporting Information Available: Contacts involving carbonyl oxygen atoms and aromatic centers observed in highresolution PDB structures and details of MD simulations, QM calculations, and AIM analysis. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes (1) Desiraju, G. R.; Steiner, T. The Weak Hydrogen Bond In Structural Chemistry and Biology; Oxford University Press: New York, 1999. (2) Steiner, T.; Koellner, G. J. Mol. Biol. 2001, 305, 535. (3) Brandl, M.; Weiss, M. S.; Jabs, A.; Suhnel, J.; Hilgenfeld, R. J. Mol. Biol. 2001, 307, 357. (4) Dougherty, D. A. Science 1996, 271, 163. (5) Ma, J. C.; Dougherty, D. A. Chem. ReV. 1997, 97, 1303. (6) Quinonero, D.; Garau, C.; Rotger, C.; Frontera, A.; Ballester, P.; Costa, A.; Deya, P. M. Angew. Chem., Int. Ed. 2002, 41, 3389. (7) Burley, S. K.; Petsko, G. A. Science 1985, 229, 23. (8) Meyer, E. A.; Castellano, R. K.; Diederich, F. Angew. Chem., Int. Ed. 2003, 42, 1210. (9) Kannan, N.; Vishveshwara, S. Protein Eng. 2000, 13, 753. (10) Zacharias, N.; Dougherty, D. A. Trends Pharmacol. Sci. 2002, 23, 281. (11) Crowley, P. B.; Golovin, A. Proteins 2005, 59, 231. (12) Kumpf, R. A.; Dougherty, D. A. Science 1993, 261, 1708. (13) Rosokha, Y. S.; Lindeman, S. V.; Rosokha, S. V.; Kochi, J. K. Angew. Chem., Int. Ed. 2004, 43, 4650. (14) Gallivan, J. P.; Dougherty, D. A. Org. Lett. 1999, 1, 103. (15) Reyes, A.; Fomina, L.; Rumsh, L.; Fomine, S. Int. J. Quantum Chem. 2005, 104, 335. (16) Korenaga, T.; Tanaka, H.; Ema, T.; Sakai, T. J. Fluorine Chem. 2003, 122, 201. (17) Sarkhel, S.; Rich, A.; Egli, M. J. Am. Chem. Soc. 2003, 125, 8998. (18) Egli, M.; Sarkhel, S. Acc. Chem. Res. 2007, 40, 197. (19) Lovell, S.; Davis, I.; Arendall, W., III; De Bakker, P.; Word, J.; Prisant, M.; Richardson, J.; Richardson, D. Proteins: Struct., Funct., Genet. 2003, 50, 437.
J. Phys. Chem. B, Vol. 111, No. 30, 2007 8683 (20) Scheiner, S.; Kar, T.; Pattanayak, J. J. Am. Chem. Soc. 2002, 124, 13257. (21) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian 03, revision B.05; Gaussian, Inc.: Wallingford, CT, 2004. (22) Dunning, T. H. J. Phys. Chem. A 2000, 104, 9062. (23) Bader, R. F. W. Atoms in Molecules. A Quantum Theory; Clarendon: Oxford, U.K., 1990. (24) Bader, R. F. W. Chem. ReV. 1991, 91, 893. (25) Garau, C.; Frontera, A.; Quinonero, D.; Ballester, P.; Costa, A.; Deya, P. M. ChemPhysChem 2003, 4, 1344. (26) Biegler-Konig, F.; Schonbohm, J.; Bayles, D. J. Comput. Chem. 2001, 22, 545. (27) Gallivan, J. P.; Dougherty, D. A. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 9459. (28) Shi, Z.; Olson, C. A.; Kallenbach, N. R. J. Am. Chem. Soc. 2002, 124, 3284. (29) Bhattacharyya, R.; Samanta, U.; Chakrabarti, P. Protein Eng. 2002, 15, 91. (30) Lindahl, E.; Hess, B.; van der Spoel, D. J. Mol. Model. 2001, 7, 306.
