Extended knobs-into-holes packing in classical and complex coiled-coil assemblies

This year marks the 50th anniversary of Crick’s seminal paper on the packing of α-helices into coiled-coil structures. The central tenet of Crick’s work is the interdigitation of side chains, which directs the helix–helix interactions; so called knobs-into-holes packing. Subsequent determinations of coiled-coil-protein sequences and structures confirmed the key features of Crick’s model and established it as a fundamental concept in structural biology. Recently, we developed a program, SOCKET, to recognise knobs-into-holes packing in protein structures, which we applied to the Protein Data Bank to compile a database of coiled-coil structures. In addition to classic structures, the database reveals 4-helix bundles and larger helical assemblies. Here, we describe how the more-complex structures can be understood by extending Crick’s principles for classic coiled coils. In the simplest case, each helix of a 2-stranded structure contributes a single seam of (core) knobs-into-holes to the helical interface. 3-, 4-, and 5-Stranded structures, however, are best considered as rings of helices with cycles of knobs-into-holes. These higher-order oligomers make additional (peripheral) knobs-into-holes that broaden the helical contacts. Combinations of core and peripheral knobs may be assigned to different sequence repeats offset within the same helix. Such multiple repeats lead to multi-faceted helices, which explain structures above dimers. For instance, coiled-coil oligomer state correlates with the offset of the different repeats along a sequence. In addition, certain multi-helix assemblies can be considered as conjoined coiled coils in which multi-faceted helices participate in more than one coiled-coil motif.     

Historical review: another 50th anniversary--new periodicities in coiled coils

In 1953, Francis Crick and Linus Pauling both proposed models of supercoiled helices (‘coiled coils’) for the structure of keratin. These were the first attempts at modelling the tertiary structure of a protein. Crick emphasized the packing mode of the side-chains (‘knobs-into-holes’), which required a periodicity of seven residues over two helical turns (7/2) and a supercoil in the opposite sense of the constituent helices. By contrast, Pauling envisaged a broader set of periodicities (4/1, 7/2, 18/5, 15/4, 11/3) and supercoils of both senses. Crick's model became canonical and the ‘heptad repeat’ essentially synonymous with coiled coils, but 50 years later new crystal structures and protein sequences show that the less common periodicities envisaged by Pauling also occur in coiled coils, adding a variant packing mode (‘knobs-to-knobs’) to the standard model. Pauling's laboratory notebooks suggest that he searched unsuccessfully for this packing mode in 1953.     

Interaction of transmembrane helices by a knobs-into-holes packing characteristic of soluble coiled coils

Membrane-embedded protein domains frequently exist as α-helical bundles, as exemplified by photosynthetic reaction centers, bacteriorhodopsin, and cytochrome C oxidase. The sidechain packing between their transmembrane helices was investigated by a nearest-neighbor analysis which identified sets of interfacial residues for each analyzed helix–helix interface. For the left-handed helix–helix pairs, the interfacial residues almost exclusively occupy positions a, d, e, or g within a heptad motif (abcdefg) which is repeated two to three times for each interacting helical surface. The connectivity between the interfacial residues of adjacent helices conforms to the knobs-into-holes type of sidechain packing known from soluble coiled coils. These results demonstrate on a quantitative basis that the geometry of sidechain packing is similar for left-handed helix–helix pairs embedded in membranes and coiled coils of soluble proteins. The transmembrane helix–helix interfaces studied are somewhat less compact and regular as compared to soluble coiled coils and tolerate all hydrophobic amino acid types to similar degrees. The results are discussed with respect to previous experimental findings which demonstrate that specific interactions between transmembrane helices are important for membrane protein folding and/or oligomerization. Proteins 31:150–159, 1998. © 1998 Wiley-Liss, Inc.     

α-Helical coiled-coil oligomerization domains in extracellular proteins

Subunit oligomerization of many proteins is mediated by α-helical coiled-coil domains. 3,4-Hydrophobic heptad repeat sequences, the characteristic feature of the coiled-coil protein folding motif, have been found in a wide variety of gene products including cytoskeletal, nuclear, muscle, cell surface, extracellular, plasma, bacterial, and viral proteins. Whereas the majority of coiled-coil structures is represented by intracellular α-helical bundles that contain two polypeptide chains, examples of extracellular coiled-coil proteins are fewer in number. Most proteins located in the extracellular space form three-stranded α-helical assemblies. Recently, five-stranded coiled coils have been identified in thrombospondins 3 and 4 in cartilage oligomeric matrix protein, and the formation of a heterotetramer has been observed in in vitro studies with the recombinant asialoglycoprotein receptor oligomerization domain. Coiled-coil domains in laminins and probably also in tenascins and thrombospondins are responsible for the formation of tissue-specific isoforms by selective oligomerization of different polypeptide chains.     

Open-and-shut cases in coiled-coil assembly: alpha-sheets and alpha-cylinders

The coiled coil is a ubiquitous protein-folding motif. It generally is accepted that coiled coils are characterized by sequence patterns known as heptad repeats. Such patterns direct the formation and assembly of amphipathic alpha-helices, the hydrophobic faces of which interface in a specific manner first proposed by Crick and termed "knobs-into-holes packing". We developed software, SOCKET, to recognize this packing in protein structures. As expected, in a trawl of the protein data bank, we found examples of canonical coiled coils with a single contiguous heptad repeat. In addition, we identified structures with multiple, overlapping heptad repeats. This observation extends Crick's original postulate: Multiple, offset heptad repeats help explain assemblies with more than two helices. Indeed, we have found that the sequence offset of the multiple heptad repeats is related to the coiled-coil oligomer state. Here we focus on one particular sequence motif in which two heptad repeats are offset by two residues. This offset sets up two hydrophobic faces separated by approximately 150 degrees -160 degrees around the alpha-helix. In turn, two different combinations of these faces are possible. Either similar or opposite faces can interface, which leads to open or closed multihelix assemblies. Accordingly, we refer to these two forms as alpha-sheets and alpha-cylinders. We illustrate these structures with our own predictions and by reference to natural variants on these designs that have recently come to light.     

MultiCoil: A program for predicting two-and three-stranded coiled coils

A new multidimensional scoring approach for identifying and distinguishing trimeric and dimeric coiled coils is implemented in the MultiCoil program. The program extends the two-stranded coiled-coil prediction program PairCoil to the identification of three-stranded coiled coils. The computations are based upon data gathered from a three-stranded coiled-coil database comprising 6,319 amino acid residues, as well as from the previously constructed two-stranded coiled-coil database. In addition to identifying coiled coils not predicted by the two-stranded database programs, MultiCoil accurately classifies the oligomerization states of known dimeric and trimeric coiled coils. Analysis of the MultiCoil scores provides insight into structural features of coiled coils, and yields estimates that 0.9% of all protein residues form three-stranded coiled coils and that 1.5% form two-stranded coiled coils. The MultiCoil program is available at http://theory.Ics.mit.edu/multicoil.     

Metal‐dependent assembly of a protein nano‐cage

Short, alpha‐helical coiled coils provide a simple, modular method to direct the assembly of proteins into higher order structures. We previously demonstrated that by genetically fusing de novo–designed coiled coils of the appropriate oligomerization state to a natural trimeric protein, we could direct the assembly of this protein into various geometrical cages. Here, we have extended this approach by appending a coiled coil designed to trimerize in response to binding divalent transition metal ions and thereby achieve metal ion‐dependent assembly of a tetrahedral protein cage. Ni²⁺, Co²⁺, Cu²⁺, and Zn²⁺ ions were evaluated, with Ni²⁺ proving the most effective at mediating protein assembly. Characterization of the assembled protein indicated that the metal ion–protein complex formed discrete globular structures of the diameter expected for a complex containing 12 copies of the protein monomer. Protein assembly could be reversed by removing metal ions with ethylenediaminetetraacetic acid or under mildly acidic conditions.     

Acid-induced polymerization of the group 5 mite allergen from Dermatophagoides pteronyssinus.

House dust mites are the most important source of indoor allergens and cause allergic diseases. Our studies here suggest that the group 5 allergen from Dermatophagoides pteronyssinus (Der p 5) is monomeric at neutral pH, but forms filaments at low pH. Circular dichroism measurements show Der p 5 is a helical protein, and the protein sequence reveals Der p 5 contains coiled-coil helices. The acid-induced filament assembly could be explained in part by the high content of charged residues (40%) in the coiled-coil structure. Interestingly, some of the known Dermatophagoides allergens also contain a heptad repeat, which could potentially form coiled coils. Therefore, coiled-coil helices may be one of the common structural motifs of mite allergens that contribute to their allergenicity.     

The Evolution and Structure Prediction of Coiled Coils across All Genomes

Coiled coils are α-helical interactions found in many natural proteins. Various sequence-based coiled-coil predictors are available, but key issues remain: oligomeric state and protein-protein interface prediction and extension to all genomes. We present SpiriCoil (http://supfam.org/SUPERFAMILY/spiricoil), which is based on a novel approach to the coiled-coil prediction problem for coiled coils that fall into known superfamilies: hundreds of hidden Markov models representing coiled-coil-containing domain families. Using whole domains gives the advantage that sequences flanking the coiled coils help. SpiriCoil performs at least as well as existing methods at detecting coiled coils and significantly advances the state of the art for oligomer state prediction. SpiriCoil has been run on over 16 million sequences, including all completely sequenced genomes (more than 1200), and a resulting Web interface supplies data downloads, alignments, scores, oligomeric state classifications, three-dimensional homology models and visualisation. This has allowed, for the first time, a genomewide analysis of coiled-coil evolution. We found that coiled coils have arisen independently de novo well over a hundred times, and these are observed in 16 different oligomeric states. Coiled coils in almost all oligomeric states were present in the last universal common ancestor of life. The vast majority of occasions that individual coiled coils have arisen de novo were before the last universal common ancestor of life; we do, however, observe scattered instances throughout subsequent evolutionary history, mostly in the formation of the eukaryote superkingdom. Coiled coils do not change their oligomeric state over evolution and did not evolve from the rearrangement of existing helices in proteins; coiled coils were forged in unison with the fold of the whole protein.     

Periodicity of amide proton exchange rates in a coiled-coil leucine zipper peptide.

The two-stranded coiled-coil motif, which includes leucine zippers, is a simple protein structure that is well suited for studies of helix-helix interactions. The interaction between helices in a coiled coil involves packing of "knobs" into "holes", as predicted by Crick in 1953 and confirmed recently by X-ray crystallography for the GCN4 leucine zipper [O'Shea, E.K., Klemm, J.D., Kim, P.S., & Alber, T. (1991) Science 254, 539]. A striking periodicity, extending over six helical turns, is observed in the rates of hydrogen-deuterium exchange for amide protons in a peptide corresponding to the leucine zipper of GCN4. Protons at the hydrophobic interface show the most protection from exchange. The NMR chemical shifts of amide protons in the helices also show a pronounced periodicity which predicts a short H-bond followed by a long H-bond every seven residues. This variation was anticipated in 1953 by Pauling and is sufficient to give rise to a local left-handed superhelical twist characteristic of coiled coils. The amide protons that lie at the base of the "hole" in the "knobs-into-holes" packing show slow amide proton exchange rates and are predicted to have short H-bond lengths. These results suggest that tertiary interactions can lead to highly localized, but substantial, differences in stability and dynamics within a secondary structure element and emphasize the dominant nature of packing interactions in determining protein structure.     

Pitch diversity in α-helical coiled coils

Two complementary methods for measuring local pitch based on heptad position in α-helical coiled coils are described and applied to six crystal structures. The results reveal a diversity of pitch values: two-stranded coiled coils appear to have pitch values near 150 Å the values for three- and four-stranded coiled coils range closer to 200 Å. The methods also provide a rapid and sensitive gauge of local coiled-coil conformation. Polar or charged residues in the apolar interface between coiled-coil helices markedly affect local pitch values, suggesting a connection between pitch uniformity and coiled-coil stability. Moreover, the identification of a skip residue (heptad frame shift) in the hemaglutinin glycoprotein of influenza virus (HA) allows interpretation of local pitch changes. These results on relatively short coiled-coil structures have relevance for the much longer fibrous proteins (many of which have skip residues) whose detailed structures are not yet established. We also show that local pitch values from molecular dynamics predictions of the GCN4 leucine zipper are in striking agreement with the high-resolution crystal structure—a result not readily discerned by direct comparison of atomic coordinates. Taken together, these methods reveal specific aspects of coiled-coil structure which may escape detection by global analyses of pitch. © 1993 Wiley-Liss, Inc.     

Predicting helix orientation for coiled-coil dimers

The alpha-helical coiled coil is a structurally simple protein oligomerization or interaction motif consisting of two or more alpha helices twisted into a supercoiled bundle. Coiled coils can differ in their stoichiometry, helix orientation, and axial alignment. Because of the near degeneracy of many of these variants, coiled coils pose a challenge to fold recognition methods for structure prediction. Whereas distinctions between some protein folds can be discriminated on the basis of hydrophobic/polar patterning or secondary structure propensities, the sequence differences that encode important details of coiled-coil structure can be subtle. This is emblematic of a larger problem in the field of protein structure and interaction prediction: that of establishing specificity between closely similar structures. We tested the behavior of different computational models on the problem of recognizing the correct orientation--parallel vs. antiparallel--of pairs of alpha helices that can form a dimeric coiled coil. For each of 131 examples of known structure, we constructed a large number of both parallel and antiparallel structural models and used these to assess the ability of five energy functions to recognize the correct fold. We also developed and tested three sequence-based approaches that make use of varying degrees of implicit structural information. The best structural methods performed similarly to the best sequence methods, correctly categorizing approximately 81% of dimers. Steric compatibility with the fold was important for some coiled coils we investigated. For many examples, the correct orientation was determined by smaller energy differences between parallel and antiparallel structures distributed over many residues and energy components. Prediction methods that used structure but incorporated varying approximations and assumptions showed quite different behaviors when used to investigate energetic contributions to orientation preference. Sequence based methods were sensitive to the choice of residue-pair interactions scored.     

Removing an interhelical salt bridge abolishes coiled-coil formation in a de novo designed peptide

Alpha-helical coiled coils represent a common protein oligomerization motif that are mainly stabilized by hydrophobic interactions occurring along their coiled-coil interface, the so-called hydrophobic seam. We have recently de novo designed and optimized a series of two-heptad repeat long coiled-coil peptides which are further stabilized by a complex network of inter- and intrahelical salt bridges. Here we have extended the de novo design of such two heptad-repeat long peptides by removing the central and most important g-e' Arg to Glu (g-e'RE) ionic interhelical interaction and replacing these residues by alanine residues. The effect of the missing interhelical ionic interaction on coiled-coil formation and stability has been analyzed by CD spectroscopy, analytical ultracentrifugation, and X-ray crystallography. We show that the peptide, while being highly alpha-helical, is no longer able to form a parallel coiled-coil structure but rather assumes an octameric globular helical assembly devoid of any coiled-coil interactions.     

Helical peptide arrays for lead identification and interaction site mapping

Libraries composed of linear and cyclic peptides cannot fully represent the higher order structures of most antigenic sites. To map the binding site of ligands or antibodies, a larger part of the three-dimensional space should be sampled. Because parallel synthesis of large arrays of peptides on hydrogels is restricted to relatively small peptides, a simple and robust homodimeric helical system was chosen for antigen presentation. First, it was established in an heterodimeric system that the 26-mer peptide could be synthesized and that the helical coiled-coil peptides interact in the hydrogel in a predictable manner. Next, libraries of homodimeric coiled coils were synthesized into which the epitope was grafted. Using dedicated helical dimeric and trimeric coiled-coil libraries, the epitopes of two anti-HIV-1 gp41 monoclonal antibodies known to interact with helical structures were mapped at high resolution. These mappings precisely reflect existing X-ray data, and the arrays can be applied to lead identification, epitope mapping, and systematic analysis of amino acid contribution to coiled-coil systems.     

A potential molecular switch in an alpha-helical coiled coil

The yeast DNA-binding protein GCN4 forms a homo-dimer through a self-complementary coiled-coil interface. In this article, we describe how such coiled-coils might be bistable and, through Molecular Dynamics computations on the GCN4 coiled coil, we show that the coiled coil can indeed switch between the two states by a pathway in which there is a progressive "flipping" of consecutive steps along the interface. We discuss the general implications of potentially bistable coiled-coil interfaces for allosteric signal-transmission mechanisms along homo-dimeric coiled coils and for the packing of helices in globular proteins.     

Multicoil2: predicting coiled coils and their oligomerization states from sequence in the twilight zone

The alpha-helical coiled coil can adopt a variety of topologies, among the most common of which are parallel and antiparallel dimers and trimers. We present Multicoil2, an algorithm that predicts both the location and oligomerization state (two versus three helices) of coiled coils in protein sequences. Multicoil2 combines the pairwise correlations of the previous Multicoil method with the flexibility of Hidden Markov Models (HMMs) in a Markov Random Field (MRF). The resulting algorithm integrates sequence features, including pairwise interactions, through multinomial logistic regression to devise an optimized scoring function for distinguishing dimer, trimer and non-coiled-coil oligomerization states; this scoring function is used to produce Markov Random Field potentials that incorporate pairwise correlations localized in sequence. Multicoil2 significantly improves both coiled-coil detection and dimer versus trimer state prediction over the original Multicoil algorithm retrained on a newly-constructed database of coiled-coil sequences. The new database, comprised of 2,105 sequences containing 124,088 residues, includes reliable structural annotations based on experimental data in the literature. Notably, the enhanced performance of Multicoil2 is evident when tested in stringent leave-family-out cross-validation on the new database, reflecting expected performance on challenging new prediction targets that have minimal sequence similarity to known coiled-coil families. The Multicoil2 program and training database are available for download from http://multicoil2.csail.mit.edu.     

Conformational transition between four and five-stranded phenylalanine zippers determined by a local packing interaction

Alpha-helical coiled coils play a crucial role in mediating specific protein-protein interactions. However, the rules and mechanisms that govern helix-helix association in coiled coils remain incompletely understood. Here we have engineered a seven heptad "Phe-zipper" protein (Phe-14) with phenylalanine residues at all 14 hydrophobic a and d positions, and generated a further variant (Phe-14(M)) in which a single core Phe residue is substituted with Met. Phe-14 forms a discrete alpha-helical pentamer in aqueous solution, while Phe-14(M) folds into a tetrameric helical structure. X-ray crystal structures reveal that in both the tetramer and the pentamer the a and d side-chains interlock in a classical knobs-into-holes packing to produce parallel coiled-coil structures enclosing large tubular cavities. However, the presence of the Met residue in the apolar interface of the tetramer markedly alters its local coiled-coil conformation and superhelical geometry. Thus, short-range interactions involving the Met side-chain serve to preferentially select for tetramer formation, either by inhibiting a nucleation step essential for pentamer folding or by abrogating an intermediate required to form the pentamer. Although specific trigger sequences have not been clearly identified in dimeric coiled coils, higher-order coiled coils, as well as other oligomeric multi-protein complexes, may require such sequences to nucleate and direct their assembly.