Giardia gene predicts a 183 kDa nucleotide-binding head-stalk protein

Intermediate filament protein evolution and protists

Metazoans evolved from a single protist lineage. While all eukaryotes share a conserved actin and tubulin-based cytoskeleton, it is commonly perceived that intermediate filaments (IFs), including lamin, vimentin or keratin among many others, are restricted to metazoans. Actin and tubulin proteins are conserved enough to be detectable across all eukaryotic genomes using standard phylogenetic methods, but IF proteins, in contrast, are notoriously difficult to identify by such means. Since the 1950s, dozens of cytoskeletal proteins in protists have been identified that seemingly do not belong to any of the IF families described for metazoans, yet, from a structural and functional perspective fit criteria that define metazoan IF proteins. Here, we briefly review IF protein discovery in metazoans and the implications this had for the definition of this protein family. We argue that the many cytoskeletal and filament-forming proteins of protists should be incorporated into a more comprehensive picture of IF evolution by aligning it with the recent identification of lamins across the phylogenetic diversity of eukaryotic supergroups. This then brings forth the question of how the diversity of IF proteins has unfolded. The evolution of IF proteins likely represents an example of convergent evolution, which, in combination with the speed with which these cytoskeletal proteins are evolving, generated their current diversity. IF proteins did not first emerge in metazoa, but in protists. Only the emergence of cytosolic IF proteins that appear to stem from a nuclear lamin is unique to animals and coincided with the emergence of true animal multicellularity.

The costa of trichomonads: A complex macromolecular cytoskeleton structure made of uncommon proteins

BACKGROUND INFORMATION

The costa is a prominent striated fibre that is found in protozoa of the Trichomonadidae family that present an undulating membrane. It is composed primarily of proteins that have not yet been explored. In this study, we used cell fractionation to obtain a highly enriched costa fraction whose structure and composition was further analysed by electron microscopy and mass spectrometry.

RESULTS

Electron microscopy of negatively stained samples revealed that the costa, which is a periodic structure with alternating electron-dense and electron-lucent bands, displays three distinct regions, named the head, neck and body. Fourier transform analysis showed that the electron-lucent bands present sub-bands with a regular pattern. An analysis of the costa fraction via one- and two-dimensional electrophoresis and liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) allowed the identification of 54 hypothetical proteins. Fourteen of those proteins were considered to be major components of the fraction.

CONCLUSIONS

The costa of T. foetus is a complex and organised cytoskeleton structure made of a large number of proteins which is assembled into filamentous structures. Some of these proteins exhibit uncharacterised domains and no function related according to gene ontology, suggesting that the costa structure may be formed by a new class of proteins that differ from those previously described in other organisms. Seven of these proteins contain prefoldin domains displaying coiled-coil regions. This propriety is shared with proteins of the striated fibres of other protozoan as well as in intermediate filaments.

SIGNIFICANCE

Our observations suggest the presence of a new class of the cytoskeleton filaments in T. foetus. We believe that our data could auxiliate in determining the specific locations of these proteins in the distinct regions that compose the costa, as well as to define the functional roles of each component. Therefore, our study will help in the better understanding of the organisation and function of this structure in unicellular organisms.

Examination of a novel head-stalk protein family in Giardia lamblia characterised by the pairing of ankyrin repeats and coiled-coil domains

The intestinal pathogen Giardia lamblia possesses several unusual organelle features, including two equivalent nuclei, no mitochondria or peroxisomes, and a developmentally regulated rough endoplasmic reticulum and Golgi. Giardia also possesses a number of complex and unique cytoskeleton structures that dictate cell shape, motility and attachment. Our investigations of cytoskeletal proteins have revealed the presence of a new protein family. Proteins in this family contain both ankyrin repeats and coiled-coil domains; although these are common protein motifs, their pairing is unique, thus establishing a new class of head-stalk proteins. Examination of the G. lamblia genome shows evidence for at least 18 genes coding for proteins with a series of ankyrin repeats followed by a lengthy coiled-coil domain and at least an additional 14 genes coding for proteins with a prominent coiled-coil domain flanked by two series of ankyrin repeats. We have examined one of these proteins, Giardia Axoneme Associated Protein (GASP-180), in detail. GASP-180 is a 180 kDa protein containing five ankyrin repeats in a 200 amino acid N-terminal domain separated by a short spacer from an approximately 1375 amino acid coiled-coil domain. Using anti-peptide antibodies raised against a unique 20 amino acid sequence found at the C-terminus, we have determined that GASP-180 is present in cytoskeleton extractions of the parasite and localises to the proximal base of the anterior flagellar axonemes. The combination of the localisation and the structural and functional motifs of GASP-180 make it a strong candidate to participate in control of flagellar activity.

The structure of alpha-helical coiled coils

alpha-Helical coiled coils are versatile protein domains, supporting a wide range of biological functions. Their fold is probably better understood than that of any other protein; indeed, uniquely among folds, their structure can be computed from a set of parametric equations. Here, we review the principles of coiled-coil structure, the determinants of their folding and stability, and the diversity of structural forms they assume.

The cytoskeleton of Giardia lamblia

Giardia lamblia is a ubiquitous intestinal pathogen of mammals. Evolutionary studies have also defined it as a member of one of the earliest diverging eukaryotic lineages that we are able to cultivate and study in the laboratory. Despite early recognition of its striking structure resembling a half pear endowed with eight flagella and a unique ventral disk, a molecular understanding of the cytoskeleton of Giardia has been slow to emerge. Perhaps most importantly, although the association of Giardia with diarrhoeal disease has been known for several hundred years, little is known of the mechanism by which Giardia exacts such a toll on its host. What is clear, however, is that the flagella and disk are essential for parasite motility and attachment to host intestinal epithelial cells. Because peristaltic flow expels intestinal contents, attachment is necessary for parasites to remain in the small intestine and cause diarrhoea, underscoring the essential role of the cytoskeleton in virulence. This review presents current day knowledge of the cytoskeleton, focusing on its role in motility and attachment. As the advent of new molecular technologies in Giardia sets the stage for a renewed focus on the cytoskeleton and its role in Giardia virulence, we discuss future research directions in cytoskeletal function and regulation.

Generalized Crick equations for modeling noncanonical coiled coils

Crick envisaged the alpha-helical coiled coil to result from systematic bending of an alpha-helix such that every seventh residue was structurally equivalent, and he derived equations for the coordinates of the backbone atoms. Crick's predictions were vindicated experimentally and coiled-coil sequences were shown to have hydrophobic residues alternately spaced 3 and 4 residues apart. Nonetheless, in some coiled coils such canonical heptad repeats are interrupted by inserts of 3 or 4 residues generating decad and hendecad motifs. The supercoiling of the coiled coils varies with the sequence pattern, being left- or right-handed in purely heptad-based or hendecad-based motifs, respectively. To model coiled coils with a mixture of motifs, we describe how Crick's equations can be modified for cases where the pitch is not constant. Using the analogy of the bending of a beam, we took the tilt angle to change linearly with distance along the major helix and the pitch of a motif to be affected by neighboring motifs depending on the rigidity of the alpha-helical strands. We tested our approach by fitting the two-, three-, and four-stranded noncanonical coiled coils of GrpE, hemagglutinin, and tetrabrachion. The backbone atoms of the model and crystal structures agreed with root mean square deviations of <1.1 A.

Investigating the tolerance of coiled-coil peptides to nonheptad sequence inserts

Coiled-coil motifs foster a wide variety of protein-protein interactions. Canonical coiled coils are based on 7-residue repeats, which guide the folding and assembly of amphipathic alpha-helices. In many cases such repeats remain unbroken for tens to hundreds of residues. However, the sequences of an increasing number of putative and characterised coiled coils digress from this pattern. We probed the consequences of nonheptad inserts using a designed leucine-zipper system. The parent peptide, SKIP0, which had four contiguous heptads, was confirmed as a parallel homodimer by circular dichroism spectroscopy and analytical ultracentrifugation. Seven daughter peptides were constructed in which 1 to 7 alanine residues were inserted between the central heptads of SKIP0. Like SKIP0, SKIP7 formed a stable helical dimer, but the other peptides were highly destabilised, with the order of dimer stability SKIP4 >> SKIP5 > SKIP6 > SKIP3 > SKIP2 > SKIP1. These results are consistent with an extended theory of coiled-coil assembly in which coiled-coil-compatible motifs are based on 3- and 4-residue spacings and most notably heptad (7-residue) and hendecad (11-residue) repeats. Thus, they help explain why in natural sequences, inserts after canonical heptad repeats most commonly comprise 4 residues. Possible biological roles for nonheptad inserts are discussed.

Biology of Giardia lamblia

Giardia lamblia is a common cause of diarrhea in humans and other mammals throughout the world. It can be distinguished from other Giardia species by light or electron microscopy. The two major genotypes of G. lamblia that infect humans are so different genetically and biologically that they may warrant separate species or subspecies designations. Trophozoites have nuclei and a well-developed cytoskeleton but lack mitochondria, peroxisomes, and the components of oxidative phosphorylation. They have an endomembrane system with at least some characteristics of the Golgi complex and encoplasmic reticulum, which becomes more extensive in encysting organisms. The primitive nature of the organelles and metabolism, as well as small-subunit rRNA phylogeny, has led to the proposal that Giardia spp. are among the most primitive eukaryotes. G. lamblia probably has a ploidy of 4 and a genome size of approximately 10 to 12 Mb divided among five chromosomes. Most genes have short 5' and 3' untranslated regions and promoter regions that are near the initiation codon. Trophozoites exhibit antigenic variation of an extensive repertoire of cysteine-rich variant-specific surface proteins. Expression is allele specific, and changes in expression from one vsp gene to another have not been associated with sequence alterations or gene rearrangements. The Giardia genome project promises to greatly increase our understanding of this interesting and enigmatic organism.

The Giardia lamblia genome

Giardia lamblia is a protozoan parasite of humans and other mammals that is thought to be one of the most primitive extant eukaryotic organisms. Although distinctly eukaryotic, it is notable for its lack of mitochondria, nucleoli, and perixosomes. It has been suggested that Giardia spp. are pre-mitochondriate organisms, but the identification of genes in G. lamblia thought to be of mitochondrial origin has generated controversy regarding that designation. Giardi lamblia trophozoites have two nuclei that are identical in all ways that have been studied. They are polyploid with at least four, and perhaps eight or more, copies of each of five chromosomes per organism and have an estimated genome complexity of 1.2x10(7)bp of DNA, and GC content of 46%. There is evidence for recombination at the telomeres of some of the chromosomes, and multiple size variants of single chromosomes have been identified within cloned isolates. However, the internal regions of the chromosomes demonstrate no evidence of recombination. For example, there is no evidence for control of vsp gene expression by DNA recombination, and no evidence for rapid mutation in the vsp genes. Single pass sequences of approximately 9% of the G. lamblia genome have already been obtained. An ongoing genome project plans to obtain approximately 95% of the genome by a random approach, as well as a complete physical map using a bacterial artificial chromosome library. The results will facilitate a better understanding of the biology of Giardia spp. as well as their phylogenetic relationship to other primitive organisms.

Analysis of striated fiber formation by recombinant SF-assemblin in vitro

The basal bodies of green flagellates are often connected to striated microtubule-associated fibers (SMAFs), which are highly ordered bundles of 2 nm filaments. SF-assemblin (33 kDa) is the principal structural subunit of the SMAFs and consists of a non-helical head domain of approximately 32 residues and an alpha-helical rod domain that shows a pronounced coiled-coil forming ability. To investigate the functional role of the head domain we expressed N-terminally truncated molecules using a cDNA coding for SF-assemblin from Chlamydomonas reinhardtii. Recombinant wild-type SF-assemblin or molecules with an N-terminal deletion of ten residues formed striated fibers with an axial repeat of 28 nm. N-terminal truncations of 19 and 29 residues yielded assembly-incompetent molecules, revealing that the head domain is necessary for the constitution of striated fibers. Further, molecules with an internal deletion of 24 residues or with duplicated segments corresponding to insertions of 29 and 58 residues were constructed. The resulting fibers had altered cross-striation patterns and axial repeats. The observed shifts in the axial repeat corresponded well to the number of inserted or deleted residues, indicating a linear coherence between molecule length and axial repeat. The heptad pattern of the rod domain of SF-assemblin is regularly interrupted by skip residues. The removal of one or two skip residues had no significant effect on the ultrastructure of the striated fibers. Substitution of skip no. 2 with alanine resulted in a modified, asymmetric cross-striation pattern, indicating a polar architecture of the striated fibers. In summary, various mutations of SF-assemblin effected the solubility of the molecules, and the axial repeat, cross-striation pattern, or overall appearance of the fibers. Thus, analysis of SF-assemblin may represent a valuable system to study the interactions involved in the polymerization of fibrous coiled-coil proteins. A model of the SMAFs based on staggered protofilaments consisting of overlapping 36 nm subunits is presented.

Heterogeneity and a coiled coil prediction of trypanosomatid-like flagellar rod proteins in Euglena

The emergent flagellum of euglenoids and trypanosomatids contained in addition to microtubules a prominent filamentous structure--the flagellar rod (paraflagellar/paraxonemal rod). Immunoblots and immunofluorescence localization using three antibodies generated against gel-isolated proteins confirmed previous studies that the Euglena flagellar rod consisted of polypeptides migrating at 66-, 69-, and 75-kD. Immunoblotting after two dimensional gel electrophoresis identified ten or more isoforms of these polypeptides. Differences in migration in acrylamide gels under nonreducing and reducing conditions suggested that the rod proteins contain intramolecular disulfide linkages. Comparative peptide mapping showed that the 66-, 69-, and 75-kD polypeptides were distinct, but related proteins, and also identified a fourth related protein migrating at 64-kD. Using antibodies against rod proteins, two overlapping cDNAs were isolated and from their sequences the cDNAs were predicted to encode 334 amino acids of the 66-kD protein; the amino acid sequence had > 65% identity to the carboxyl-terminus of the trypanosomatid flagellar rod proteins. Secondary structural prediction suggested that flagellar rod proteins contain an extended segmented coiled coil stalk and two nonhelical heads. Coiled coil appeared to be an important structural motif in the construction of flagellar rod filaments.

Coiled-coil assembly by peptides with non-heptad sequence motifs

BACKGROUND

The seven-residue heptad repeat is the accepted hallmark of coiled coils. In extended filamentous proteins, however, contiguous patterns of heptads are often disrupted by 'skips' and 'stammers'. The structural consequences and roles of these digressions are not understood.

RESULTS

In a cytoskeleton protein from Giardia lamblia, heptads flank eleven-residue units (hendecads) to give a 7-11-7 motif that dominates the sequence. Synthetic peptides made to the consensus sequence of this motif fold in solution to fully helical, parallel dimers. Both the sequence pattern and these experimental data are consistent with the coiled-coil model. We note that breaks in other extended coiled coils can also be reconciled by hendecad insertions.

CONCLUSIONS

The heptad paradigm for the coiled coil must be expanded to include hendecads. As different combinations of heptads and hendecads will give different overall sequence motifs, we propose that these provide a mechanism to promote cognate protein pairings during the folding of extended coiled coils in the cell.

Heptad breaks in alpha-helical coiled coils: stutters and stammers

The discontinuities found in heptad repeats of alpha-helical coiled-coil proteins have been characterized. A survey of 40 alpha-fibrous proteins reveals that only two classes of heptad breaks are prevalent: the stutter, corresponding to a deletion of three residues, and the newly identified "stammer," corresponding to a deletion of four residues. This restriction on the variety of insertions/deletions encountered gives support to a unifying structural model, where different degrees of supercoiling accommodate the observed breaks. Stutters in the hemagglutinin coiled-coil region have previously been shown to produce an underwinding of the supercoil, and we show here how, in other cases, stammers would lead to overwinding. An analysis of main-chain structure also indicates that the mannose-binding protein, as well as hemagglutinin, contains an underwound coiled-coil region. In contrast to knobs-into-holes packing, these models give rise to non-close-packed cores at the sites of the heptad phase shifts. We suggest that such non-close-packed cores may function to terminate certain coiled-coil regions, and may also account for the flexibility observed in such long alpha-fibrous molecules as myosin. The local underwinding or overwinding caused by these specific breaks in the heptad repeat has a global effect on the structure and can modify both the assembly of the protein and its interaction properties.