Alpha-helical coiled coils and bundles: how to design an alpha-helical protein

Phasing the conformational unit of spectrin

Many proteins contain a repetitive sequence motif, which implies that they contain a repetitive structural motif. Spectrin and the related proteins dystrophin and alpha-actinin consist largely of repeated motifs of 100-120 residues. But the repeating motif is degenerate and it has been difficult to define the boundaries of the repeating sequence unit or its corresponding structural unit. We have determined at which residues the structural units that correspond to spectrin's repeating 106-amino acid motifs begin and end. Drosophila alpha-spectrin cDNAs were expressed in bacteria to show that single segments (106 amino acids) and pairs of segments encoded by selected regions of spectrin cDNA can fold into stable conformations whose biophysical and biochemical properties are similar to those of native spectrin. Because such folding was critically dependent on the phasing of the expressed sequence with respect to the apparent boundaries of the repeating motifs, our data provide experimental evidence that relates the boundaries of the folded, conformational unit to the chemical sequence of repeating motifs.

Fibrinogen structure in projection at 18 A resolution. Electron density by co-ordinated cryo-electron microscopy and X-ray crystallography

Electron microscope images of frozen-hydrated crystals of a proteolytically modified fibrinogen show excellent preservation of the structure. An electron density map of the key centric projection of the crystal at 18 A resolution has been obtained by combining the phases derived from cryo-electron microscopy with X-ray amplitudes. Simulation methods developed in earlier studies have been used to interpret the map. In contrast to the earlier images, the map allows us to visualize the coiled-coil region of the molecule and possible substructure in the beta domains. The map also shows that there is a marked difference in density in the two regions corresponding to the molecular ends where the gamma domains interact. A possible interpretation of this finding is provided by assuming substructure in the gamma domains and the breaking of molecular symmetry where these domains interact. Some additional constraints useful for the determination of the three-dimensional structure were obtained from cryo-electron micrographs of a perpendicular view at 25 A resolution. Implications of this working model for the molecular length and contacts in the filaments in both the crystal and fibrin are described. The data used here will be valuable as a starting point for obtaining the three-dimensional structure.

Do interhelical side chain-backbone hydrogen bonds participate in formation of leucine zipper coiled coils?

The leucine zipper proteins are a group of transcriptional regulators that dimerize to form a DNA binding domain. It has been proposed that this dimerization results from the hydrophobic association of the alpha-helices of two leucine zipper monomers into a coiled coil. We propose a model for a coiled coil based on a periodic hydrophobic-hydrophilic amino acid motif found in the leucine zipper regions of 11 transcriptional regulatory proteins. This model predicts the symmetrical formation of secondary hydrogen bonds between the polar side chains of one helix and the peptide carbonyls of the opposite chain, supplementing the interactions between hydrophobic side chains. Physical modeling (CPK) and in vacuo molecular mechanics calculations of the stability of the GCN4 leucine zipper coiled coil configured in accordance with this model demonstrate a greater stability for this conformer than for a conformer configured according to a current hydrophobic model. Molecular dynamics simulations show similar stability of the two models in vacuo but a higher stability of the hydrophobic model in water.

X-ray structure of the GCN4 leucine zipper, a two-stranded, parallel coiled coil

The x-ray crystal structure of a peptide corresponding to the leucine zipper of the yeast transcriptional activator GCN4 has been determined at 1.8 angstrom resolution. The peptide forms a parallel, two-stranded coiled coil of alpha helices packed as in the "knobs-into-holes" model proposed by Crick in 1953. Contacts between the helices include ion pairs and an extensive hydrophobic interface that contains a distinctive hydrogen bond. The conserved leucines, like the residues in the alternate hydrophobic repeat, make side-to-side interactions (as in a handshake) in every other layer of the dimer interface. The crystal structure of the GCN4 leucine zipper suggests a key role for the leucine repeat, but also shows how other features of the coiled coil contribute to dimer formation.

Kinetics of folding and unfolding of alpha alpha-tropomyosin and of nonpolymerizable alpha alpha-tropomyosin

Stopped flow CD (SFCD) kinetic studies of self-assembly of coiled coils of rabbit alpha alpha-tropomyosin and of nonpolymerizable alpha alpha-tropomyosin (NPTm) are reported. The protein was denatured in 6 M urea buffer, then renatured by 10-fold dilution into benign saline buffer. Folding was monitored by SFCD in the backbone region (222 nm). Protein chains are shown to be totally unfolded (and separated in the reduced species) in the initial denaturing medium and fully folded as two-chain coiled coils in the final benign medium. In all cases of folding in benign buffer of totally unfolded chains, two phases were found in the folding process: a fast phase (less than 0.04 s, the SFCD dead time), in which an intermediate state with about 70% of the equilibrium ellipticity forms; followed by a slower, observable phase that completes the folding. The slow phase is first order (k-1 = 1.6 s at 20 degrees C), signifying that chain association for reduced samples occurs in the fast phase. In contrast, folding in benign buffer from an initial state with 70% of the equilibrium ellipticity is all fast, suggesting that the folding intermediate is not an equilibrium species. Cross-linking at Cys-190 increases the helix content of the fast-formed intermediate state to about 85% of the equilibrium value, but leaves the rate constant of the slow phase unchanged. In NPTm, which does not form high aggregates at low ionic strength, the rate of the observable phase is almost independent of ionic strength in the range of approximately 0.15-0.6 M, but is reduced one to two orders of magnitude by further reduction to 0.026 M. In folding from totally unfolded chains, the rate is reduced less than one order of magnitude by changing the final state to about 50% folded. In contrast to folding, unfolding of alpha alpha-tropomyosin from the native state is all fast.

Construction, purification, and characterization of a hybrid protein comprising the DNA binding domain of the LexA repressor and the Jun leucine zipper: a circular dichroism and mutagenesis study

An increasing number of eukaryotic transcription factors interacting specifically with DNA comprise a dimerization motif called the "leucine zipper". These leucine zipper proteins form homodimers and/or heterodimers with another protein containing a leucine zipper motif. The leucine zipper of the oncoprotein Jun is particular in that Jun may form homodimers as well as heterodimers with the oncoprotein Fos, which are however more stable than the Jun-Jun homodimers. Leucine zipper dimerization is thought to occur through a coiled-coil arrangement of parallel alpha-helices, but the rules governing the specificity of homo- and/or heterodimerization are still largely unknown. To address this question in the case of the Jun leucine zipper, we constructed a fusion protein containing the amino-terminal DNA binding domain of the LexA repressor from Escherichia coli fused to the Jun leucine zipper. This hybrid protein (LexA-JunZip) is stable in E. coli and confers much tighter repression in vivo than the DNA binding domain of LexA alone. DNA binding competition experiments with synthetic Jun and Fos leucine zipper peptides in vitro showed that the leucine zipper mediated dimerization of LexA-JunZip is essential for DNA binding of the fusion protein. The purified LexA-JunZip protein dimerizes in vitro with a dimerization constant of 2 x 10(7) M-1 at 5 degrees C. Dimerization is very sensitive to temperature, since the dimerization constant drops at 20 degrees C to 2 x 10(6) M-1 and at 30 degrees C to only 3 x 10(5) M-1.(ABSTRACT TRUNCATED AT 250 WORDS)

Asymmetric inclusion by de novo designed proteins: fluorescence probe studies on amphiphilic alpha-helix bundles

The inclusion feature and supersecondary structure of the de novo designed proteins which are constructed with several amphiphilic alpha-helices and flexible linkage parts were investigated with fluorescence probes. Five types of small proteins (or peptides) have been designed, which are composed of 2, 3, 4, 4, and 6 helices, respectively, and are linked with only linear junctions except for one of 4-helix proteins. All of these proteins have inclusion ability for hydrophobic fluorophores. Further, by the analysis of fluorescence polarization anisotropy, it was suggested that these proteins include guest molecules in compact helix bundles constructed with about 4 helices. Asymmetric inclusion of both monomer and stacked dimer of acridine orange derivatives was found by means of induced circular dichroism except for the 4-helix protein with cross-junction. The chirality of the included dimer proved to be in accordance with the chiral sense of alpha-helical coiled-coil. The 6-helix protein has especially high efficiency in inclusion for any fluorophores examined in this study and brings about a significant blue-shift of maximal emission for 8-anilino-1-naphthalenesulfonate.

Common structural motifs in proteins of the extracellular matrix

Proteins of the extracellular matrix are composed of many structurally and often functionally different autonomous domains which frequently occur as modular units in several different extracellular matrix proteins, but also in proteins of different origin. Some domains serve related assembly functions in different proteins but for domains involved in cell attachment and other cellular activities only a few generalizations are possible.

DNA-induced increase in the alpha-helical content of C/EBP and GCN4

Leucine zipper proteins comprise a recently identified class of DNA binding proteins that contain a bipartite structural motif consisting of a "leucine zipper" dimerization domain and a segment rich in basic residues responsible for DNA interaction. Protein fragments encompassing the zipper plus basic region domains (bZip) have previously been used to determine the conformational and dynamic properties of this motif. In the absence of DNA, the coiled-coil portion is alpha-helical and dimeric, whereas the basic region is flexible and partially disordered. Addition of DNA containing a specific recognition sequence induces a fully helical conformation in the basic regions of these fragments. However, the question remained whether the same conformational change would be observed in native bZip proteins where the basic regions might be stabilized in an alpha-helical conformation even in the absence of DNA, through interactions with portions of the protein not included in the bZip motif. We have now examined the DNA-induced conformational transition for an intact bZip protein, GCN4, and for the bZip fragment of C/EBP with two enhancers that are differentially symmetric. Our results are consistent with the induced helical fork model wherein the basic regions are largely flexible in the absence of DNA and become fully helical in the presence of the specific DNA recognition sequence.

C-terminal structure and mobility of rabbit skeletal muscle light meromyosin as studied by one- and two-dimensional 1H NMR spectroscopy and X-ray small-angle scattering

Intact rabbit myosin and two different C-terminal fragments of rabbit muscle light meromyosin (LMM) expressed in Escherichia coli, LMM-30, and LMM-30C', were studied by 1H NMR spectroscopy. X-ray small-angle scattering shows that at high ionic strength two polypeptide chains of LMM-30 (which consists of the C-terminal 262 amino acids of myosin heavy chain) or LMM-30C' (which corresponds to LMM-30 but lacks the last 17 residues) assemble to form an alpha-helical coiled-coil as it is found also in myosin. The last 12 C-terminal residues of one polypeptide chain of LMM-30 and the last 9 C-terminal residues of the other chain are very mobile. The last 8 residues of the two strands are equivalent from the NMR point of view and unfolded; the valine residues in position 255 in the two strands are not equivalent, suggesting an interaction between the two strands, Ser-252, Arg-253, and Asp-254 are completely immobilized in one of the polypeptide strands and partly mobile in the other. Essentially the same pattern is observed in intact myosin. In spite of the large molecular weights of LMM-30 and LMM-30C', it is possible to resolve almost all aromatic residues and to determine the pK values of all the 4 tyrosine and of 9 (out of 10) histidine residues. The tyrosine residues in the two strands are equivalent in the two polypeptide chains and both have a pK of 10.5. The pK values of the histidine residues vary between 5.7 and 7.0.

Identification of FAP locus genes from chromosome 5q21

Recent studies suggest that one or more genes on chromosome 5q21 are important for the development of colorectal cancers, particularly those associated with familial adenomatous polyposis (FAP). To facilitate the identification of genes from this locus, a portion of the region that is tightly linked to FAP was cloned. Six contiguous stretches of sequence (contigs) containing approximately 5.5 Mb of DNA were isolated. Subclones from these contigs were used to identify and position six genes, all of which were expressed in normal colonic mucosa. Two of these genes (APC and MCC) are likely to contribute to colorectal tumorigenesis. The MCC gene had previously been identified by virtue of its mutation in human colorectal tumors. The APC gene was identified in a contig initiated from the MCC gene and was found to encode an unusually large protein. These two closely spaced genes encode proteins predicted to contain coiled-coil regions. Both genes were also expressed in a wide variety of tissues. Further studies of MCC and APC and their potential interaction should prove useful for understanding colorectal neoplasia.

Sequence and structural analysis of surface protein antigen I/II (SpaA) of Streptococcus sobrinus

Streptococcal antigen I/II or the surface protein antigen A (SpaA) of Streptococcus sobrinus is an adhesin which mediates binding of the organism to tooth surfaces. The complete sequence of the gene which encodes SpaA has been determined. The gene consists of 4,584 bp and encodes a protein of 1,528 amino acid residues. The deduced amino acid sequence shows extensive homology with those of the cell surface adhesins from Streptococcus mutans serotypes c and f and from Streptococcus sanguis. Structural analysis of the N-terminal region (residues 50 to 550), which is rich in alanine and includes four tandem repeats of an 82-residue sequence, suggests that it adopts an alpha-helical coiled-coil conformation. Cell surface hydrophobicity may be associated with this region. The C-terminal region is more conserved and includes two tandem repeats of a 39-residue proline-rich sequence. A further proline-rich sequence in this region is predicted to span the cell wall. Although a hydrophobic sequence is present in the C-terminal region, it appears to be too short to span the cell membrane. Anchoring of SpaA in the cell membrane may therefore require some form of posttranslational modification or association with another membrane protein.

Heptad motifs within the distal subdomain of the coiled-coil rod region of M protein from rheumatic fever and nephritis associated serotypes of group A streptococci are distinct from each other: nucleotide sequence of the M57 gene and relation of the deduced amino acid sequence to other M proteins

Streptococcal M protein, a dimeric alpha helical coiled-coil molecule, is an antigenically variable virulence factor on the surface of the bacteria. Our recent conformational analysis of the complete sequence of the M6 protein led us to propose a basic model for the M protein consisting of an extended central coiled-coil rod domain flanked by a variable N-terminal and a conserved C-terminal end domains. The central coiled-coil rod domain of M protein, which constitutes the major part of the M molecule, is made up of repeating heptads of the generalized sequence a-b-c-d-e-f-g, wherein "a" and "d" are predominantly apolar residues. Based on the differences in the heptad pattern of apolar residues and internal sequence homology, the central coiled-coil rod domain of M protein could be further divided into three subdomains I, II, and III. The streptococcal sequelae rheumatic fever (RF) and acute glomerulonephritis (AGN) have been known to be associated with distinct serotypes. Consistent with this, we observed that the AGN associated M49 protein exhibits a heptad motif that is distinct from the RF associated M5 and M6 proteins. Asn and Leu predominated in the "a" and "d" positions, respectively, in subdomain I of the M5 and M6 proteins, whereas apolar residues predominated in both these positions in the M49 protein. To establish whether the heptad motif of M49 is unique to this protein, or is a general characteristic of nephritis-associated serotypes, the amino acid sequence of M57, another nephritis-associated serotype, has now been examined. The gene encoding M57 was amplified by PCR, cloned into pUC19 vector, and sequenced. The C-terminal half of M57 is highly homologous to other M proteins (conserved region). In contrast, its N-terminal half (variable region) revealed no significant homology with any of the M proteins. Heptad periodicity analysis of the M57 sequence revealed that the basic design principles, consisting of distinct domains observed in the M6 protein, are also conserved in the M57 molecule. However, the heptad motif within the coiled-coil subdomain I of M57 was distinct from M5 and M6 but similar to M49. Similar analyses of the heptad characteristics within the reported sequences of M1, M12, and M24 proteins further confirmed the conservation of the overall architectural design of sequentially distinct M proteins.(ABSTRACT TRUNCATED AT 400 WORDS)

Point mutations that lock Salmonella typhimurium flagellar filaments in the straight right-handed and left-handed forms and their relation to filament superhelicity

We have determined the nucleotide sequence of two mutant and the parent fliC genes, encoding the protein flagellin (serotype i), of Salmonella typhimurium. The flagellar filaments of the two mutants, SJW1655 and SJW1660, are locked in the straight-right-handed (R) and straight-left-handed (L) conformations, respectively. Their normal, wild-type, parent strain is SJW1103. These mutant strains differ from the wild-type by only one base-pair: the mutation of SJW1655 occurs at nucleotide 1346 in the flagellin gene, changing a C.G pair to T.A (alanine 449 to valine). The mutation of SJW1660 occurs at nucleotide 1277, changing a G.C pair to C.G (glycine 426 to alanine). The resulting amino acid substitutions are near the C terminus predicted to form an alpha-helical coiled coil. The region contains six heptad repeats. Similar alpha-helical segments (three and four repeats long) are present near the N terminus. Alignment of the 17 flagellin sequences available to date confirms the generality of these segments. The mutations are within that portion of the sequence assigned, by proteolytic cleavage, to the middle flagellin domain whose length corresponds to the six heptad repeats found in the sequence (approximately 50 A). We have shown that these mutations are the sole cause of the straight phenotype by replacing the mutated segments with a wild-type one and restoring both superhelicity and motility.

Rotavirus YM gene 4: analysis of its deduced amino acid sequence and prediction of the secondary structure of the VP4 protein

We have determined the complete nucleotide sequence of the VP4 gene of porcine rotavirus YM. It is 2,362 nucleotides long, with a single open reading frame coding for a protein of 776 amino acids. A phylogenetic tree was derived from the deduced YM VP4 amino acid sequence and 18 other available VP4 sequences of rotavirus strains belonging to different serotypes and isolated from different animal species. In this tree, VP4 proteins were grouped by the hosts that the corresponding viruses infect rather than by the serotypes they belong to, suggesting that this protein is involved in the host specificity of the viruses. In an attempt to predict the secondary structure of the VP4 protein, we selected the more divergent VP4 sequences and made a secondary structure analysis of each protein. In spite of variations within the individual structures predicted, there was a general structural pattern which suggested the existence of at least two different domains. One, comprising the amino-terminal 63% of the protein, is predicted to be a possible globular domain rich in beta-strands alternated with turns and coils. The second domain, represented by the remaining, carboxy-terminal part of VP4, is rich in long stretches of alpha-helix, one of which, 63 amino acids long, has heptad repeats resembling those found in proteins known to form alpha-helical coiled-coils. The predicted secondary structure correlates well with the available data on the protein accessibility delineated by immunological and biochemical findings and with the spike structure of the protein, which has been determined by cryoelectron microscopy.

The kinetics of chain exchange in two-chain coiled coils: alpha alpha- and beta beta-tropomyosin

Measurements are presented on the time course of chain exchange among two-chain alpha-helical coiled coils of rabbit tropomyosin. All experiments are in a regime (temperature, protein concentration) in which coiled-coil dimers are the predominant species. Self-exchange in alpha alpha-tropomyosin was investigated by mixing alpha alpha species with alpha* alpha*, the asterisk designating an alpha-chain whose lone sulfhydryl (C190) has been blocked by carboxyamidomethylation. The overall process alpha alpha + alpha* alpha* in equilibrium with 2 alpha alpha* is followed by measurement of the fraction (h) of alpha alpha* species as a function of time. Similarly, self-exchange in beta beta-tropomyosin is examined by measurements of the overall process: beta beta + beta* beta* in equilibrium with 2 beta beta*, in which beta* signifies a beta-chain blocked at both sulfhydryls (C36 and C190). The observed time course for both chains is well fit by the first-order equation: h (t) = h (infinity) (1-e-k1t), with h (infinity) congruent to 0.5. This long-time limit is as expected for self-exchange, and agrees with experiments that attain equilibrium after slow cooling of thermally dissociated and unfolded chains. The simplest consonant mechanism is chain exchange by rate-limiting dissociation of dimers followed by random reassociation. Kinetic analysis shows k1 to be the rate constant for the chain dissociation step, a quantity not previously measured for any coiled coil. This rate constant for beta beta species is about an order of magnitude greater than for alpha alpha. In both, the activation enthalpy and entropy are very large, suggesting that activation to an extensively (greater than 50%) unfolded species necessarily precedes dissociation. Experiments are also reported for overall processes: alpha alpha + beta* beta* in equilibrium with 2 alpha beta* and alpha* alpha* + beta beta in equilibrium with 2 alpha* beta. Results are independent of which chain is blocked. Again h (infinity) congruent to 0.5, in agreement with equilibrium experiments, and the time course is first order. The rate constants and activation parameters are intermediate between those for self-exchange.

Glycine loops in proteins: their occurrence in certain intermediate filament chains, loricrins and single-stranded RNA binding proteins

Quasi-repetitive, glycine-rich peptide sequences are widespread in at least three distinct families of proteins: the keratins and other intermediate filament proteins, including nuclear lamins; loricrins, which are major envelope components of terminally differentiated epithelial cells; and single-stranded RNA binding proteins. We propose that such sequences comprise a new structural motif termed the 'glycine loop'. The defining characteristics of glycine loop sequences are: (1) they have the form x(y)n, where x is usually an aromatic or occasionally a long-chain aliphatic residue; y is usually glycine but may include polar residues such as serine, asparagine, arginine, cysteine, and rarely other residues; and the value of n is highly variable, ranging from 1 to 35 in examples identified to date. (2) Glycine-loop-containing domains are thought to form when at least two and to date, as many as 18, such quasi-repeats are configured in tandem, so that the entire domain in a protein may be 50-150 residues long. (3) The average value of n, the pattern of residues found in the x position and the non-glycine substitutions in the y position appear to be characteristic of a given glycine loop containing domain, whereas the actual number of repeats is less constrained. (4) Glycine loop sequences display a high degree of evolutionary sequence variability and even allelic variations among different individuals of the same vertebrate species. (5) Glycine loop sequences are expected to be highly flexible, but possess little other regular secondary structure.(ABSTRACT TRUNCATED AT 250 WORDS)

Predicting coiled coils from protein sequences

The probability that a residue in a protein is part of a coiled-coil structure was assessed by comparison of its flanking sequences with sequences of known coiled-coil proteins. This method was used to delineate coiled-coil domains in otherwise globular proteins, such as the leucine zipper domains in transcriptional regulators, and to predict regions of discontinuity within coiled-coil structures, such as the hinge region in myosin. More than 200 proteins that probably have coiled-coil domains were identified in GenBank, including alpha- and beta-tubulins, flagellins, G protein beta subunits, some bacterial transfer RNA synthetases, and members of the heat shock protein (Hsp70) family.

Immunoglobulin variable-region-like domains of diverse sequence within the major histocompatibility complex of the chicken

The highly polymorphic B-G antigens are considered to be part of the major histocompatibility complex (MHC) of the chicken, the B system of histocompatibility, because they are encoded in a family of genes tightly linked with the genes encoding MHC class I and class II antigens. To better understand these unusual MHC antigens, full-length B-G cDNA clones were isolated from B21 embryonic erythroid cell cDNA library, restriction-mapped, and sequenced. Five transcript types were identified. Analysis of the deduced amino acid sequences suggests that the B-G polypeptides are composed of single extracellular domains that resemble immunoglobulin domains of the variable-region (V) type, single membrane-spanning domains typical of integral membrane proteins, and long cytoplasmic tails. Sequence diversity among the five transcript types was found in all domains, notably including the B-G immunoglobulin V-like domains. The cytoplasmic tails of the B-G antigens are made up entirely of units of seven amino acid residues (heptads) that are typical of an alpha-helical coiled-coil conformation. The heptads vary in number and sequence between the different transcripts. The presence within B-G polypeptides of polymorphic immunoglobulin V-like domains warrants further investigations to determine the degree and nature of variability within this domain in these unusual MHC antigens.

Chromatographic and physical studies of tropomyosin in aqueous-organic media at low pH

Non-cross-linked and disulfide-cross-linked two-chain molecules comprising the alpha and/or beta chains of rabbit skeletal tropomyosin were studied by electrophoretic, chromatographic and physical methods. Elution order on C4 reversed-phase high-performance liquid chromatography depends markedly on the number and position of the cross-links. In the C4 reversed-phase elution medium, cross-linked and non-cross-linked species are greater than 85% helical by circular dichroism, but the non-cross-linked elute later from high-performance size-exclusion chromatography (G4000) and have molecular mass of 31,000-41,000 dalton by equilibrium ultracentrifugation. The data suggest that in the C4 reversed-phase high-performance liquid chromatography elution medium non-cross-linked tropomyosin exists as amphipathic single-chain alpha-helices.

A 43-kilodalton pneumococcal surface protein, PspA: isolation, protective abilities, and structural analysis of the amino-terminal sequence

PspA is an antigenically variable surface protein of Streptococcus pneumoniae that appears to be essential for full pneumococcal virulence. In addition, monoclonal antibodies to PspA protect mice against infection with specific strains of pneumococci virulent for mice. In this study, we have isolated the 43-kDa N-terminal half of the native 84-kDa PspA and determined the sequence of the first 45 amino acids. This sequence, the first obtained for a pneumococcal surface protein, is consistent with that of an amphiphatic coiled-coil alpha helix with a 7-residue periodicity common to fibrous proteins such as tropomyosin and streptococcal M protein. The 7-residue periodicity begins with residue 8 and extends throughout the remaining sequence for nearly 11 turns of the helix. Mice immunized with this purified PspA segment were protected from fatal pneumococcal challenge, thus demonstrating that those PspA epitopes eliciting protection were present in the N-terminal half of the molecule.

Characterization of Neurospora CPC1, a bZIP DNA-binding protein that does not require aligned heptad leucines for dimerization

CPC1 is the transcriptional activator of amino acid biosynthetic genes of Neurospora crassa. CPC1 function in vivo was abolished upon deletion of segments of cpc-1 corresponding to the presumed transcription activation domain, the DNA-binding and dimerization domains, or a 52-residue connector segment of CPC1. A truncated CPC1 polypeptide containing only the carboxy-terminal 57-residue segment of CPC1 was sufficient to form homodimers that bound DNA. However, deletion of the segment of cpc-1 corresponding to the connector segment in the full-length CPC1 polypeptide abolished DNA binding. Removal of a segment of cpc-1 corresponding to the GIn-rich region of CPC1 reduced in vivo function only slightly. The homologous transcription activator of Saccharomyces cerevisiae, GCN4, did not substitute for CPC1 in N. crassa. Chimeric CPC1-GCN4 polypeptides that contained the GCN4 transcriptional activation domain or the domain of GCN4 that corresponds to the essential 52-residue connector segment of CPC1, functioned with reduced efficiency. However, a chimeric polypeptide containing the GCN4 DNA-binding and dimerization domains in place of those of CPC1 functioned essentially as well as wild-type CPC1. The basic and dimerization domains of CPC1 were characterized by introducing deletions or site-directed amino acid replacements. The basic region was required for DNA binding but not for dimerization. CPC1 has a short dimerization domain containing heptad residues Leu-1, Leu-2, Trp-3, and His-4. When Val was substituted for Leu-1 or Leu-2, CPC1 was fully active, but when Val replaced Trp-3, dimerization and DNA binding were prevented. DNA band shift analyses with CPC1 heterodimers demonstrated that CPC1 does not require aligned heptad leucine residues for dimerization. Replacement of two charged residues located between Leu-1 and Leu-2 of CPC1 abolished dimerization and DNA binding.

Molecular analysis of protein assembly in muscle development

The challenge presented by myofibril assembly in striated muscle is to understand the molecular mechanisms by which its protein components are arranged at each level of organization. Recent advances in the genetics and cell biology of muscle development have shown that in vivo assembly of the myofilaments requires a complex array of structural and associated proteins and that organization of whole sarcomeres occurs initially at the cell membrane. These studies have been complemented by in vitro analyses of the renaturation, polymerization, and three-dimensional structure of the purified proteins.

Novel myosin heavy chain encoded by murine dilute coat colour locus

Hundreds of murine dilute mutations have been identified and analysed, making dilute one of the best genetically characterized of all mammalian loci. The recessive dilute (d) coat colour mutation carried by many inbred strains of mice produces a lightening of coat colour, caused by an abnormal adendritic melanocyte morphology that results in an uneven release of pigment granules into the developing hair shaft. Most dilute alleles (dilute-lethal) also produce a neurological defect, characterized by convulsions and opisthotonus, apparent at 8-10 days of age and continuing until the death of the animal at 2-3 weeks of age. The discovery that the original dilute allele (now termed dilute-viral or dV) is the result of the integration of an ecotropic murine leukaemia provirus has allowed the cloning of genomic DNA and in this study complementary DNA, from the dilute locus. The predicted dilute amino-acid sequence indicates that dilute encodes a novel type of myosin heavy chain, with a tail, or C-terminal, region that has elements of both type II (alpha-helical coiled-coil) and type I (non-coiled-coil) myosin heavy chains. Dilute transcripts are differentially expressed in both embryonic and adult tissues and are very abundant in neurons of the central nervous system, cephalic ganglia, and spinal ganglia. These results suggest an important role for the dilute gene product in the elaboration, maintenance, or function of cellular processes of melanocytes and neurons.

Kinetics of self-assembly of alpha alpha-tropomyosin coiled coils from unfolded chains

The two-chain coiled-coil structural motif is found in fibrous muscle proteins and leucine zippers. Unfolding/refolding studies abound, but none establishes the time scale or mechanism of structural assembly from separated, unfolded chains. Stopped-flow circular dichroism studies of such refolding of alpha-tropomyosin chains are reported here. The backbone spectral region (222 nm) reveals a fast phase (less than 0.04 s), yielding an intermediate possessing approximately 70% of the equilibrium helix content. A subsequent slow phase is first order [k-1 (20 degrees C) = 1.67 s, Ea = 12.7 kcal.mol-1 (1 kcal = 4.18 kJ)], so dimerization is fast. The same rate constant characterizes folding in the Cys-190 crosslinked chains, so the intermediate has parallel and nearly registered chains. The tyrosine spectral region (280 nm) reveals only a fast phase, so these six chain sites are native in the intermediate.

Characterization of Neurospora CPC1, a bZIP DNA-binding protein that does not require aligned heptad leucines for dimerization

CPC1 is the transcriptional activator of amino acid biosynthetic genes of Neurospora crassa. CPC1 function in vivo was abolished upon deletion of segments of cpc-1 corresponding to the presumed transcription activation domain, the DNA-binding and dimerization domains, or a 52-residue connector segment of CPC1. A truncated CPC1 polypeptide containing only the carboxy-terminal 57-residue segment of CPC1 was sufficient to form homodimers that bound DNA. However, deletion of the segment of cpc-1 corresponding to the connector segment in the full-length CPC1 polypeptide abolished DNA binding. Removal of a segment of cpc-1 corresponding to the GIn-rich region of CPC1 reduced in vivo function only slightly. The homologous transcription activator of Saccharomyces cerevisiae, GCN4, did not substitute for CPC1 in N. crassa. Chimeric CPC1-GCN4 polypeptides that contained the GCN4 transcriptional activation domain or the domain of GCN4 that corresponds to the essential 52-residue connector segment of CPC1, functioned with reduced efficiency. However, a chimeric polypeptide containing the GCN4 DNA-binding and dimerization domains in place of those of CPC1 functioned essentially as well as wild-type CPC1. The basic and dimerization domains of CPC1 were characterized by introducing deletions or site-directed amino acid replacements. The basic region was required for DNA binding but not for dimerization. CPC1 has a short dimerization domain containing heptad residues Leu-1, Leu-2, Trp-3, and His-4. When Val was substituted for Leu-1 or Leu-2, CPC1 was fully active, but when Val replaced Trp-3, dimerization and DNA binding were prevented. DNA band shift analyses with CPC1 heterodimers demonstrated that CPC1 does not require aligned heptad leucine residues for dimerization. Replacement of two charged residues located between Leu-1 and Leu-2 of CPC1 abolished dimerization and DNA binding.

Three-stranded alpha-fibrous proteins: the heptad repeat and its implications for structure

Amino acid sequence data have been collected for the coiled-coil rod domains of three-stranded alpha-fibrous proteins--fibrinogen, laminin, tenascin, macrophage scavenger receptor protein and the leg fibre protein from bacteriophage. Such domains are characterized by a heptad substructure in which apolar residues occur alternately three and four residues apart. The distribution of residues in each position of the heptad has been analysed, and the results compared with those obtained for the two-stranded alpha-fibrous proteins, which include the intermediate filament and myosin families. Distinctions can be drawn between the sequences in two- and three-stranded coiled-coil structures and these provide criteria that will prove useful in predicting secondary and tertiary structure purely from sequence data.

Sequence requirements for coiled-coils: analysis with lambda repressor-GCN4 leucine zipper fusions

A genetic system was developed in Escherichia coli to study leucine zippers with the amino-terminal domain of bacteriophage lambda repressor as a reporter for dimerization. This system was used to analyze the importance of the amino acid side chains at eight positions that form the hydrophobic interface of the leucine zipper dimer from the yeast transcriptional activator, GCN4. When single amino acid substitutions were analyzed, most functional variants contained hydrophobic residues at the dimer interface, while most nonfunctional sequence variants contained strongly polar or helix-breaking residues. In multiple randomization experiments, however, many combinations of hydrophobic residues were found to be nonfunctional, and leucines in the heptad repeat were shown to have a special function in leucine zipper dimerization.

Differentiation between two biologically distinct classes of group A streptococci by limited substitutions of amino acids within the shared region of M protein-like molecules

Group A streptococci can be categorized into two classes (I and II) based on immunodeterminants contained within a surface-exposed, conserved region (C repeat domain) of the major virulence factor, M protein. Previous studies have shown that several biological properties correlate strongly with streptococcal class, and thus, there is a strong impetus to precisely define the antigenic epitopes unique to class I and II M proteins. Using synthetic peptides, the binding sites of two class I-specific mAbs were mapped to distinct epitopes within the C repeat region of type 6 M protein (class I). A class II M protein-like gene (type 2) was cloned and sequenced, and the predicted amino acid sequence was compared for homology to class I and II molecules, whose sequences were previously reported. For a given C repeat block 35 amino acid residues in length, 20 residue positions were conserved among all sequences analyzed. Of the 15 variable amino acid positions, only four were class specific, and three of the four positions were localized in the area to which the class I-specific mAbs bound. The predicted secondary structures of class I and II C repeat blocks reveals that they are alpha-helical, except for a single area of disruption. In the class I molecules, the area of disruption corresponds to the class I-specific mAb binding sites. Importantly, the predicted conformational characteristics of this disruption differs for class I and II molecules. The data suggest that only limited changes in amino acid residues differentiate between class I and II molecules in the C repeat region. Therefore, selective (biological) pressures may have contributed to the evolution of these two classes of molecules.

Evidence for coiled-coil dimer formation by an Epstein-Barr virus transactivator that lacks a heptad repeat of leucine residues

Two regions of the Epstein-Barr virus (EBV) BZLF1 gene product, ZEBRA, share sequence homology with c-Fos, one of which corresponds to the DNA binding domain of c-Fos. ZEBRA does not, however, contain the heptad repeat of leucines present in the dimerization domains of leucine zipper proteins. Here it is shown that ZEBRA binds its recognition sites as a homodimer and that the region adjacent to the basic DNA binding domain is essential for dimerization. This region contains a 4-3 repeat of predominantly hydrophobic residues, which is precisely in register with the hydrophobic heptad repeat present in the leucine zipper proteins with respect to the basic DNA binding domain. A mutational analysis of ZEBRA supports a model for dimerization involving a coiled-coil interaction. These results indicate that a heptad repeat of leucines is not a structural requirement for formation of coiled-coil dimers by transcription factors.

Molecular cloning and expression of a hexameric Drosophila heat shock factor subject to negative regulation

We report the cloning of the transcriptional activator of heat shock genes, HSF, from Drosophila. The predicted sequence of Drosophila HSF protein is surprisingly divergent from that of yeast HSF, except in regions important for DNA binding and oligomerization. A segment of the DNA binding domain of HSF bears an intriguing similarity to the putative DNA recognition helix of bacterial sigma factors, while the oligomerization domain contains an unusual arrangement of conserved hydrophobic heptad repeats. Drosophila HSF produced in E. coli under nonshock conditions forms a hexamer that binds specifically to DNA with high affinity and activates transcription from a heat shock promoter in vitro. In contrast, when HSF is expressed in Xenopus oocytes, maximal DNA binding affinity is observed only after heat shock induction. These results suggest that Drosophila HSF has an intrinsic affinity for DNA, which is repressed under nonshock conditions in vivo.

A thermodynamic scale for the helix-forming tendencies of the commonly occurring amino acids

Amino acids have distinct conformational preferences that influence the stabilities of protein secondary and tertiary structures. The relative thermodynamic stabilities of each of the 20 commonly occurring amino acids in the alpha-helical versus random coil states have been determined through the design of a peptide that forms a noncovalent alpha-helical dimer, which is in equilibrium with a randomly coiled monomeric state. The alpha helices in the dimer contain a single solvent-exposed site that is surrounded by small, neutral amino acid side chains. Each of the commonly occurring amino acids was substituted into this guest site, and the resulting equilibrium constants for the monomer-dimer equilibrium were determined to provide a list of free energy difference (delta delta G degree) values.

Size polymorphism of chicken major histocompatibility complex-encoded B-G molecules is due to length variation in the cytoplasmic heptad repeat region

B-G antigens are cell-surface molecules encoded by a highly polymorphic multigene family located in the chicken major histocompatibility complex (MHC). Rabbit antisera to B-G molecules immunoprecipitate 3-6 bands from iodinated erythrocytes by sodium dodecyl sulfate (SDS) gels under reducing conditions. These are all B-G molecules because they all map to the B-G region of the chicken MHC in congenic and recombinant chickens, most are directly recognized by the antisera, most form disulfide-linked dimers, and none bear N-linked carbohydrate. Both apparent homodimers and heterodimers are found, which bear intrachain disulfide bonds. All 3-6 bands have different mobilities in SDS gels between different haplotypes, ranging from 30 to 55 kDa. This size polymorphism is not affected by glycosidase treatment or addition of protease inhibitors. Partial proteolysis of cell surface-iodinated B-G molecules generates extremely similar patterns of spots, both within and between haplotypes. These surface-iodinated peptides bear either interchain or intrachain disulfide bonds. Additional peptides are generated by proteolysis of B-G molecules iodinated after isolation. Thus, it appears that the extracellular regions of these molecules are very similar and that the length polymorphism is due to variations in the cytoplasmic regions. Inspection of the cDNA-derived protein sequence in this region shows many heptad repeats, which may allow variation in length by step deletion and alternative splicing. The repeats indicate an alpha-helical coiled-coil structure, which could form an interaction between subunits of the dimer or with the cytoskeleton or both.

Structural features in the heptad substructure and longer range repeats of two-stranded alpha-fibrous proteins

Considerable sequence data have been collected from the intermediate filament proteins and other alpha-fibrous proteins including myosin, tropomyosin, paramyosin, desmoplakin and M-protein. The data show that there is a clear preference for some amino acids to occur in specific positions within the heptad substructure that characterizes the sequences which form the coiled-coil rod domain in this class of proteins. The results also indicate that although there are major similarities between the various proteins there are also key differences. In all cases, however, significant regularities in the linear disposition of the acidic and the basic residues in the coiled-coil segments can be related to modes of chain and molecular aggregation. In particular a clear trend has been observed which relates the mode of molecular aggregation to the number of interchain ionic interactions per heptad pair.

Parallel zippers formed by alpha-helical peptide columns in crystals of Boc-Aib-Glu(OBzl)-Leu-Aib-Ala-Leu-Aib-Ala-Lys(Z)-Aib-OMe

The crystal structure of the decapeptide Boc-Aib-Glu(OBzl)-Leu-Aib-Ala-Leu-Aib-Ala-Lys(Z)-Aib-OMe (where Aib is alpha-aminoisobutyryl, Boc is t-butoxycarbonyl, OBzl is benzyl ester, and Z is benzyloxycarbonyl) illustrates a parallel zipper arrangement of interacting helical peptide columns. Head-to-tail NH...OC hydrogen bonding extends the alpha-helices formed by the decapeptide into long columns in the crystal. An additional NH...OC hydrogen bond in the head-to-tail region, between the extended side chains of Glu(OBzl), residue 2 in one molecule, and Lys(Z), residue 9 in another molecule, forms a "double tooth" on the side of the column. These double teeth are repeated regularly on the helical columns with spaces of six residues between them (approximately 10 A). The double teeth on a pair of parallel columns (all carbonyl groups pointed in the same direction) interdigitate in a zipper motif. All contacts in the zipper portion are of the van der Waals type. The peptide, with formula C66H103N11O17.H2O, crystallizes in space group P2(1)2(1)2(1) with a = 10.677(4) A, b = 16.452(6) A, and c = 43.779(13) A; overall agreement R = 10.2% for 3527 observed reflections (magnitude of /F0/ greater than 3 sigma); resolution 0.9 A.

Haemopoietic receptors and helical cytokines

A bewitching interplay of proteins, variously clothed as chemical messengers and cellular receptors, control the pace of growth and the course of progressive differentiation in blood cell types. The messengers are lymphokines, interleukins, colony-stimulating factors, growth hormones and interferons: generically, the cytokines. The second components of the regulatory pairs are membrane-spanning receptor proteins: these molecules transduce the specific binding of cognate cytokines into a mitogenic cellular response. In this article, Fernando Bazan discusses a provocative structural model for cytokine-receptor interactions which, if correct, will alter perceptions of the evolutionary design of the haemopoietic system.

A second class of synthetase structure revealed by X-ray analysis of Escherichia coli seryl-tRNA synthetase at 2.5 A

The three-dimensional crystal structure of seryl-transfer RNA synthetase from Escherichia coli, refined at 2.5 A resolution, is described. It has an N-terminal domain that forms an antiparallel alpha helical coiled-coil, stretching 60 A out into the solvent and stabilized by interhelical hydrophobic interactions and an active-site alpha-beta domain based around a seven-stranded antiparallel beta sheet. Unlike the three other known synthetase structures, the enzyme contains no classical nucleotide-binding fold, and is the first representative of a second class of aminoacyl-tRNA synthetase structures.

Type I macrophage scavenger receptor contains alpha-helical and collagen-like coiled coils

The macrophage scavenger receptor is a trimeric membrane glycoprotein with unusual ligand-binding properties which has been implicated in the development of atherosclerosis. The trimeric structure of the bovine type I scavenger receptor, deduced by complementary DNA cloning, contains three extracellular C-terminal cysteine-rich domains connected to the transmembrane domain by a long fibrous stalk. This stalk structure, composed of an alpha-helical coiled coil and a collagen-like triple helix, has not previously been observed in an integral membrane protein.

Alpha-helix to random-coil transitions of two-chain coiled coils: the use of physical models in treating thermal denaturation equilibria of isolated subsequences of alpha alpha-tropomyosin

Two extant models of thermal folding/unfolding equilibria in two-chain, alpha-helical coiled coils are tested by comparison with experimental results on excised, isolated subsequences of rabbit alpha alpha-tropomyosin (Tm). These substances are designated iTmj where i and j are, respectively, the residue numbers (in the 284-residue parent chain) of the N- and C-terminal residues of the subsequence. One model postulates that a coiled coil consists of segments, each denaturing in an all-or-none manner, like small globular proteins. Thus this model yields a small number of populated molecular species. In an extant calorimetry study of 11Tm127 and of 190Tm284, each required only two all-or-none-segments, and their enthalpies and transition temperatures were assigned. These assignments are shown here to yield the concentration of all molecular species, and therefore the helix content, as a function of temperature. Such calculations for 190Tm284 are in tolerable agreement with CD experiments, but those for 11Tm127 are in gross disagreement. Thus, either the model itself or the calorimetric assignment is faculty. In the second model, all conformational states are counted and weighted, as in the Zimm-Bragg theory for single-chain polypeptides. This theory has been extended (by Skolnick) to two-chain coiled coils and is here used to fit CD data for 11Tm127, 142Tm281, and 190Tm284. The fit is tolerable for 11Tm127, good for 142Tm281, and quantitative for 190Tm284. Thus this comparison does not falsify this second model. The helix-helix interaction free energy, obtainable from the fit, shows nonadditivity when isolated subsequences are compared with the parent. This suggests that removal of a region from a long coiled coil allows energetically substantial adjustments in side-chain packing in the helix-helix interface. Thus, the helix-helix interaction in long coiled coils is characteristic of a global free energy minimum and not just of the regional constellation of side chains.