Hydrogen bonding plays a key role in structure and function of proteins including features such as protein folding, local architecture, protein-ligand recognition, enzymatic activity, protein hydration and molecular dynamics. These classical hydrogen bonds are well established and have been studied in detail (Baker and Hubbard, 1984; Jeffrey and Saenger, 1991). However, a set of somewhat weaker interactions has also been recognized to play an important role in protein structure and stability (Desiraju and Steiner, 1999). This set includes N-H...π, O-H...π, C-H...O and C-H...π. The overall stabilization energy of these interactions is much smaller (not more than a few Kcal/mol), but is still signficant.

C-H...X interaction

In recent years it has been established that a C-H group can be a hydrogen bond donor. Although considered weak in nature, the bond C-H...X (X: O, π) termed as C-H interaction is known to be distrubuted widely among protein structures. These interactions have shown to be of greater importance in proteins, and polypeptides, protein-protein, protein-ligand and drug-binding interactions.

        

    C-H...O interaction                                       C-H...π interaction

C-H...O interaction

C-H...O interactions form 20-25% of the total number of hydrogen bonds constituting the second most important group (Weiss, 2001). Important early work indicated a role for C-H...O interactions in protein structures. Steiner and Saenger (1993) reported the presence of C-H...O bonds completing the coordination of the buried water molecules in papain. Derewenda et al. 1994 presented evidence of functional significance in serine hydrolases of a hydrogen bond involving the C-H group of the active site His and a neighboring main-chain carbonyl oxygen. The past studies have revealed that C^α-H...O hydrogen bonds occur almost ubiquitously in beta-sheets (Fabiola et al. 1997), but they also appear to occur frequently in alpha-helical proteins (Chakrabarti and Chakrabarti, 1998).

Identification of C^α-H...O interactions

The C-H...O interactions in the dataset have been identified using the program HBexplore (Lindauer et al., 1996) that requires structural information in the Protein Data Bank (PDB) format (Bernstein et al., 1977). The set I criteria with geometrical parameters (distance H...O < 2.5Α, C^α...O < 3.9Α and C^α-H...O > 90^o) have been used. The program identifies all potential hydrogen bonds according to geometrical criteria and generates an output. The output lists the name, position number of donor and acceptor residues along with geometrical parameters and subgroup of the identified H-bond set e.g. backbone-backbone bond.

C-H...π interaction

Hydrogen bonds, C-H...π with π-acceptor constitute yet another considerable fraction. Steiner and Koellner, 2001 described hydrogen bonds in proteins involving aromatic acceptors, and Brandl et al. exhaustively surveyed the occurrence of interactions involving all possible C-H groups (C^α, C_aliphatic-H and C_aromatic-H) as donors and all possible side chain π systems as acceptors. The cases in which C-H...π interactions have been described in proteins include the formation of complexes of proteins with ligands or cofactors such as the heme group (Nishio et al. 1998) and design of serine proteases inhibitors (Shimohigashi et al. 1996, 1999). Previous studies have also shown that C-H...π interactions are even responsible for the stabilization of structural elements such as alpha or 3₁₀ helices or non-proline cis peptide bonds (Brandl et al.,2001) .

Identification of C^α-H...π interactions

The C-H...π interactions in the dataset have been identified using the web server NCI server that is based purely on geometric criteria (Babu, 2003) . The default parameters (C^α...O=C <= 3.8Α H^α...O=C <=3.3Α, C^α-H...O >=120^o and H^α...O=C >= 90^o) have been used. The output consists of name, position of interacting residues and the observed values for each of the parameters.