Journal of Molecular Biology
Volume 315, Issue 2, 11 January 2002, Pages 171-182
Journal home page for Journal of Molecular Biology

Regular article
A novel main-chain anion-binding site in proteins: the nest. A particular combination of φ,ψ values in successive residues gives rise to anion-binding sites that occur commonly and are found often at functionally important regions1

https://doi.org/10.1006/jmbi.2001.5227Get rights and content

Abstract

Main-chain conformations where one amino acid residue can be described as γR (or αR) and an adjacent one as γL (or αL) mostly result in the three main-chain NH groups (of the two residues and the one following) forming a depression that can accommodate an atom with a whole or partial negative charge. We propose the name nest for this feature. The negatively charged atom, when present, is also stabilized by hydrogen-bonding with the NH groups. In an average protein, 8 % of residues are involved in a nest. The anion, or partially negatively charged atom, that often occupies the nest may be a main-chain carbonyl oxygen atom as in the paperclip, also called the Schellman loop, and the oxyanion hole of serine proteases. It can be a phosphate group, as in the P-loop superfamily that binds ATP and GTP. Overlapping, compound, nests are observed often, as in the P-loop, which has five successive NH groups that bind the β phosphate group of nucleotide triphosphate. The longest compound nests are found surrounding cysteine-bound [2Fe2S] and [4Fe4S] iron-sulfur centers, which are also anionic; nests may encourage binding of the more reduced forms. The nest is a novel feature in the sense of not having been described as a unique motif with anion-binding potential before, although some of the situations where it occurs are familiar.

Introduction

In spite of much work1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22 describing small, mainly hydrogen-bonded, motifs of under ten residues in proteins, it may seem surprising to come across common and functionally useful protein features of this sort recurring in a variety of situations that have remained essentially unrecognised. Individual types are known and even familiar, but the structural similarities and relationships between them have not, as far as we can tell, been described hitherto.

Here, we regard carbonyl groups as anionic because of their partial charge. In relation to anion binding sites, carbonyl oxygen atoms belonging to amides behave similarly to those of carboxylates. A well-known situation where main-chain NH groups form an anion-binding site is at the N terminus of an α-helix. Because their NH groups all point to the C terminus, most α-helices have two or three free NH groups not hydrogen-bonded to main-chain CO groups and these often form anion-binding sites of various sorts. They may bind carbonyl groups for example,14, 21 or phosphate groups.23 They sometimes occur on their own without forming a proper α-helix (residues 27-29 of phospholipases A2,24 bound to the carbonyl group of the conserved aspartate 43, form one), identifiable by the αR or γRdihedral angles of a minimum of two successive residues. However, these motifs are not the primary concern of the present work.

We point to another feature where main-chain NH groups form a potential anion-binding site. It is characterized by alternating residues with γR(or αR) and γL(or αL) main-chain dihedral angles in either order. A consequence of two such adjacent residues is often the formation of an atom-sized concavity with their two main-chain NH groups, plus the NH of the amino acid following, pointing inwards, ideal for an anion-binding site, as seen in Figure 1(a). The first and third NH groups are well positioned for hydrogen bonding to an atom in the cavity generated, while the middle NH sometimes also does so though it points more to the side. We propose the word nest as a name for this three NH motif.

Just as α-helices in native proteins can be right-handed or, rarely, left-handed, so nests may exist as one of two enantiomers, this form of enantiomerism being restricted to the main-chain atoms. We use the terminology RL and LR to refer to them. RL is where the first residue is αR or γR and the second αL or γL. If the residues are the other way round the nest is of type LR. Both are found, although RL nests are more common. However, the oxyanion hole in serine proteases is an LR nest.

This survey reveals that compound nests of the form RLR, and tandem nests of the form RLLR, are found in proteins and are prominent in certain functional regions. In the compound type, which is commoner, the two overlapping nests each have a potential binding site for an anionic atom facing in about the same direction, such that they form a wider concavity that may accommodate anionic groups rather than just single atoms. Long compound nests we have found occur in the P-loops of nucleotide triphosphate-binding proteins (LRLR, with five consecutive NH groups facing into the concavity) and in the loops that bind [2Fe2S] or [4Fe4S] in iron-sulfur proteins (RLRLR in some ferredoxins, with six such NH groups).

Section snippets

Why nests constitute a novel class of motif

We define nests from the φ,ψ angles of successive residues given in Methods. Essentially, one set is in the γR or αR region and one is γL or αL. The distribution of residues in proteins in the Ramachandran plot is skewed towards the αR or αL, rather than the γR or γL, regions both in terms of the distribution observed and energetic conformational favorability. Yet, in the motifs found by these criteria, there are relatively few αRαL or αLαRcombinations. This appears to be related to the

Conclusion

The nest, the name we give to the depression formed by the main-chain NH groups of three consecutive amino acid residues in a particular conformation, is a common means by which main-chain atoms of proteins bind anions for various purposes. The paperclip/Schellman loop is one of the commonest features exhibiting a nest; it can be regarded as a means of stabilizing the carbonyl groups at α-helical C termini. Such nests bind two oxygen atoms, whereas others bind only one. Nests occur in two

Methods

The primary database, used as the basis for the work described first (Figure 1, Figure 2, Figure 3, Figure 4, Table 1, Table 2, Table 3, Table 4), is a set of 67 protein structures, influenced by the ‘PDB_SELECT’ list at ftp://ftp.embl-heidelberg.de/pub/databases/pdb_select.51 All have less than 25 % sequence identity with each other, are refined by X-ray diffraction at a resolution of less than 1.5 Å, and have an R-factor less than 20 %. The PDB codes, with the subunit letter added when

References (57)

  • D Dreusicke et al.

    The glycine-rich loop of adenylate kinase forms a giant anion hole

    FEBS Letters

    (1986)
  • E.J Milner-White

    The N-terminal domain of MDM2 resembles calmodulin and its relatives

    J. Mol. Biol.

    (1999)
  • S Iwata et al.

    Structure of a water-soluble fragment of the ‘Rieske’ iron-sulfur protein of the bovine heart mitochondrial cytochrome bc1 complex

    Structure

    (1996)
  • C.W Carter

    New stereochemical analogies between iron sulfur electron transfer proteins

    J. Biol. Chem.

    (1977)
  • D.G Vassylyev et al.

    Crystal structures of ribonuclease F1 from Fusarium moniliforme in its free form and complexed with 2′GMP

    J. Mol. Biol.

    (1993)
  • F Gliubich et al.

    Active site structural features of chemically modified forms of rhodanese

    J. Biol. Chem.

    (1996)
  • E.J Milner-White

    Beta-bulges within loops as recurring motifs of protein structure

    Biochim. Biophys. Acta

    (1987)
  • J.A Cowan et al.

    Structure-function correlations in high potential iron proteins

    Advan. Inorg. Chem.

    (1998)
  • E.J Milner-White et al.

    Loops, bulges, turns and hairpins in proteins

    Trends Biochem. Sci.

    (1987)
  • R Preissner et al.

    Occurrence of bifurcated three-centre hydrogen bonds in proteins

    FEBS Letters

    (1991)
  • D.F Stickle et al.

    Hydrogen bonding in globular proteins

    J. Mol. Biol.

    (1992)
  • J.M Hemmingsen et al.

    The tyrosine cornera feature of most greek key β-barrel proteins

    Protein Sci.

    (1994)
  • D Gandini et al.

    Patterns in ionisable side-chain interactions in protein structures

    Proteins: Struct. Funct. Genet.

    (1996)
  • L.E Donate et al.

    Conformational analysis and clustering of short and medium-sized loops connecting regular secondary structures

    Protein Sci.

    (1996)
  • A.J Doig et al.

    Structures of N termini of helices in proteins

    Protein Sci.

    (1997)
  • L Shimoni et al.

    Hydrogen bonding motifs of protein side-chains

    Protein Sci.

    (1995)
  • K Gunasekaran et al.

    β-Hairpins in proteins revisitedlessons for ne novo design

    Protein Eng.

    (1997)
  • G.A Jeffrey

    An Introduction to Hydrogen Bonding

    (1997)
  • Cited by (147)

    • Determination of amino acids that favour the α<inf>L</inf> region using Ramachandran propensity plots. Implications for α-sheet as the possible amyloid intermediate

      2021, Journal of Structural Biology
      Citation Excerpt :

      In studies of β-turn occupancy of the αL region Wilmot and Thornton (1990) and Efimov (1993) distinguished two distinct subregions centred on (60°,30°) and (90°,0°); we denote these αL and γL respectively. Nests have two successive residues in αRγL (RL) or γLαR (LR) conformations; RL nests make up 80% of such motifs in proteins (Watson and Milner-White, 2002b), while 20% are LR. Strands of α-sheet (α-strands), on the other hand, consist of alternating αR/αL residues, that are, approximately, adjacent right-handed/left-handed α-helical conformations.

    • New insights into the effect of sludge proteins on the hydrophilic/hydrophobic properties that improve sludge dewaterability during anaerobic digestion

      2020, Water Research
      Citation Excerpt :

      While cation binding which interact selectively and non-covalently with cations, charged atoms or groups of atoms with a net positive charge. The removal of anion binding protein indicated that the sludge dewaterability was improved during the AD process, because the anion binding proteins have more free amino groups, which is the main reason for the sludge surface hydrophilicity (Habeeb, 1966), the amino groups and hydrogen bonding, more energy is required to remove (White and Watson, 2002). And amino groups are hydrophilic, sludge surface presents more hydrophobic because the reduction of amino groups.

    View all citing articles on Scopus
    1

    Edited by J. Thornton

    View full text