Stranded in isolation: structural role of isolated extended strands in proteins

Protein Engineering vol. 16 no. 5 pp. 331±339, 2003 DOI: 10.1093/protein/gzg046 Stranded in isolation: structural role of isolated extended strands in proteins Narayanan Eswar1, C.Ramakrishnan and N.Srinivasan2 Molecular Biophysics Unit, Indian Institute of Science, Bangalore 560 012, India 1Present address: Departments of Biopharmaceutical Sciences and Pharmaceutical Chemistry, University of California, San Francisco, San Francisco, CA 94143-2240, USA whom correspondence should be addressed. E-mail: Reasons for the formation of extended-strands (E-strands) in proteins are often associated with the formation of bsheets. However E-strands, not part of b-sheets, commonly occur in proteins. This raises questions about the structural role and stability of such isolated E-strands. Using a dataset of 250 largely non-homologous and high-resolution Ê ) crystal structures of proteins, we have identi®ed (<2 A 518 isolated E-strands from 187 proteins. The two most distinguishing features of isolated E-strands from bstrands in b-sheets are the high preponderance of prolyl residues occuring in isolated E-strands and their high exposure to the surroundings. Removal of regions with polyproline conformation from the dataset did not signi®cantly reduce the propensity of prolyl residues to occur in isolated E-strands. Isolated E-strands are often characterized by their main-chain amide and carbonyl groups involved in hydrogen bonding with polar side chains or water. They are often ¯anked by irregular loop structures and are less well conserved, than b-sheet forming bstrands, among homologous protein structures. It is suggested that isolated b-strands have many characteristics of loop segments but with repetitive (f,y) values falling within the b-region of the Ramachandran map. Keywords: b-sheet/b-strand/extended strand/hydrogen bonding/protein structures Introduction The requirement of polar groups in proteins to be satis®ed by hydrogen bonding can be viewed as a director of protein folding (Rose and Wolfenden, 1993). As most of the amino acid residues in the interior of protein structures are known to lack polar side chains (Chothia, 1976; Miller et al., 1987), it is conceivable that most of the polar groups at the interior are situated at the backbone of the polypeptide chain. These polar groups of the polypeptide backbone (NH and C=O groups) are known often to be satis®ed by virtue of the formation of helical and b-sheet structures in proteins (Baker and Hubbard, 1984; Stickle et al., 1992). Formation of characteristic hydrogen bonding patterns involving the amide and carbonyl groups of the polypeptide main chain is an essential feature of the formation of a-helices, b-sheets and b-turns in proteins (Pauling and Corey, 1951; Pauling et al., 1951; Venkatachalam, 1968). Indeed, an important driving factor Protein Engineering 16(5), ã Oxford University Press 331 2To for the formation of a-helix in proteins is suggested to be the formation of intra-segment hydrogen bonding (Presta and Rose, 1988). Deviation from the characteristic hydrogen bonding patterns in a-helices and b-sheets is known to result in distortions in these structures (Richardson et al., 1978; Barlow and Thornton, 1988). These regions of distortion are often found to be solvated. For example, the kink produced by a proline residue in the middle of an a-helix and the existence of a b-bulge in b-sheets are well known. The amino acid residue preferences and van der Waals stabilizing interactions are also characteristics of a-helices and b-strands in proteins (Street and Mayo, 1999). The conformational entropy for the rotation of side chains is suggested to be a key feature in the preference or otherwise of an amino acid type to occur in a-helix or b-sheet form (Presta and Rose, 1988; Creamer and Rose, 1992; 1994; Stapley and Doig, 1997). For example, interactions between the side chains in positions i and i + 3 (and i + 4) in a-helices (Creamer and Rose, 1995) and interactions between side chains across b-strands involved in the formation of a b-sheet are known to contribute to the stabilization of these structures (Lifson and Sander, 1980; Otzen and Fersht, 1995; Smith and Regan, 1995; Wouters and Curmi, 1995). The b-sheet is generally considered as a `secondary structure' although it is known to be distinct from the other kinds of regular secondary structures. The distinction stems from the fact that it requires spatially neighbouring regions of the protein, in extended conformation, to become aligned to form the characteristic inter-strand hydrogen bonds. However, it may be inappropriate to refer to the b-strand as a secondary structure as, unlike other kinds of secondary structure, there are no intra-segment hydrogen bonds. Often, it is tempting to associate the role of formation of a main-chain region in the extended conformation (extended strands or E-strands) with that of b-sheets. In this paper, we draw attention to the regions of proteins in extended conformation that are not involved in the formation of a b-sheet. As the description of an extended strand does not involve the hydrogen bonding of amide and carbonyl groups of the backbone, unless involved in the formation of a b-sheet, the role of such extended structures in proteins is puzzling. Also, as these E-strands are not participating in the formation of b-sheet there is no possibility of inter-strand interaction between nonpolar residues like the one ®rst observed by Lifson and Sander (1980). We have surveyed a large number of known protein structures and found that such isolated extended strands commonly occur in proteins and share characteristics of loops and b-sheets in proteins. These E-strands are distinct from the polyproline type II extended conformation whose occurrence in globular protein structures has been extensively studied (Soman and Ramakrishnan, 1983; Adzhubei et al., 1987a±c; Ananthanarayanan et al., 1987; Adzhubei and Sternberg, 1993). The polyproline type II conformation is N.Eswar, C.Ramakrishnan and N.Srinivasan Materials and Methods Dataset used Ê ) and nonA dataset of 250 highly resolved (resolution <2.0 A homologous protein structures derived from the Protein Data Bank (PDB) (Bernstein et al., 1977; Berman et al., 2000) was used for the analysis. In the case of proteins with identical or very similar polypeptide chains, only one of them was considered. The chain used in such cases is shown as a ®fth character in the complete list of PDB codes of the proteins used as follows: 1aan, 1aazA, 1abe, 1abk, 1acf, 1acx, 1afgA, 1ahc, 1ak3A, 1alc, 1ald, 1alkA, 1amp, 1ankA, 1aozA, 1apmE, 1arb, 1arp, 1ars, 1ast, 1bbhA, 1bbpA, 1bgc, 1bgh, 1bmdA,1brsD, 1bsaA, 1byb, 1cbn, 1ccr, 1cewI, 1cgt, 1chmA, 1cmbA, 1cot, 1cpcA, 1cpcB, 1cpn, 1cseE, 1cse I, 1csh, 1ctf, 1cus, 1ddt, 1dfnA, 1dmb, 1dri, 1dsbA, 1eca, 1esl, 1ezm, 1fas, 1fdn, 1fgvH, 1®aA, 1fkf, 1¯p, 1¯v, 1fna, 1frrA, 1fus, 1fxl, 1fxd, 1gd1O, 1gia, 1gky, 1glqA, 1glt, 1gog, 1gox, 1gp1A, 1gp (...truncated)