Molecular Evolution and Structure of -Actinin
http://www.100md.com
分子生物学进展 2004年第6期
Department of Biochemistry, Ume University, SE-901-87 Ume, Sweden
E-mail: lars.backman@chem.umu.se.
Abstract
The N-terminal actin-binding domain of -actinin is connected to the C-terminal EF-hands by a rod domain. Because of its ability to form dimers, -actinin can cross-link actin filaments in muscle cells as well as in nonmuscle cells. In the prototypic -actinins, the rod domain contains four triple helical bundles, or so-called spectrin repeats. We have found some atypical -actinins in early diverging organisms, such as protozoa and yeast, where the rod domain contains one and two spectrin repeats, respectively. This implies that the four repeats present in modern -actinins arose after two consecutive intragenic duplications from an -actinin with a single repeat. Further, the evolutionary gene tree of -actinins shows that the appearance of four distinct -actinin isoforms may have occurred after the vertebrate-invertebrate split. The topology of the tree lends support to the hypothesis that two rounds (2R) of genome duplication occurred early in the vertebrate radiation. The phylogeny also considers these atypical isoforms as the most basal to -actinins of vertebrates and other eukaryotes. The analysis also positioned -actinin of the fungi Encephalitozoo cuniculi close to the protozoa, supporting the suggestion that microsporidia are early eukaryotes. Because -actinin is considered the basal member of the spectrin family, our studies will improve the understanding of the origin and evolution of this superfamily.
Key Words: -actinin ? phylogeny ? evolution ? spectrin superfamily ? spectrin repeat
Introduction
-Actinin is a ubiquitous actin cross-linker belonging to the spectrin superfamily (Blanchard, Ohanian, and Critchley 1989; Dubreuil 1991). It has been identified in most eukaryotic organisms, from human (Beggs et al. 1992; Mills et al. 2001) and mouse (Mills et al. 2001) to fly (Drosophila melanogaster) and worm (Caenorhabditis elegans) (Fyrberg et al. 1990; Barstead, Kleiman, and Waterston 1991). Although fission yeast (Schizosaccharomyces pombe) appears to express an -actinin (or at least an -actinin–like protein) baker's yeast (Saccharomyces cerevisiae) does not (Wu, Bahler, and Pringle 2001).
The structure of the -actinin has been determined in a great detail. Depending on the source, this elongated protein has a molecular mass of 93 to 103 kDa (Blanchard, Ohanian, and Critchley 1989). The actin-binding site is located at the N-terminus, which comprises two calponin homology domains, where only the most N-terminal one binds actin (Norwood et al. 2000). At the C-terminus, there are two EF-hands, where the second EF-hand is crucial for calcium binding (Witke et al. 1993; Janssen et al. 1996). These two domains are connected by the rod domain. This domain is formed by triple-helical repeats or so-called spectrin repeats, where each repeat contains approximately 106 amino acid residues (Djinovic-Carugo et al. 1999). The rod domain of -actinin is important for dimerization (Djinovic-Carugo et al. 1999; Ylanne et al. 2001). There are usually four spectrin repeats in the rod domain, but -actinin from some organisms appears to have only one or two repeats.
Because the rod domain determines the distance between cross-linked actin filaments and serves as an interaction site for receptors and adaptor proteins, the number of repeats has implications for actin bundling as well as for cytoskeleton organization (Djinovic-Carugo et al. 1999; Ylanne et al. 2001; Djinovic-Carugo et al. 2002).
-Actinins are classified into four different main isoforms (Dixson, Forstner, and Garcia 2003). Humans and probably other vertebrates express all four isoforms, whereas invertebrates and protista seem to express only one isoform. The muscle isoforms, -actinin 2 and of -actinin 3 are localized to the Z-disc of the sarcomeres (Blanchard, Ohanian, and Critchley 1989; Mills et al. 2001). These two isoforms are insensitive to calcium because their EF-hands are nonfunctional (Beggs et al. 1992). In contrast, the nonmuscle isoforms, -actinin 1 and of -actinin 4 are sensitive to calcium (Tang, Taylor, and Taylor 2001). -Actinin 1 and of -actinin 4 can be found at the leading edge of motile cells, at cell adhesion sites and focal contacts, and along actin stress fibers in migrating cells (Barstead, Kleiman, and Waterston 1991).
To better understand the structural and evolutionary relationships between the -actinin isoforms, we have characterized -actinin from the urochordate Ciona intestinalis, as well as several other primitive eukaryotes. Our results demonstrate that the general structural pattern of -actinin is not applicable to the most primitive eukaryotes, as these isoforms are shorter because of fewer spectrin repeats in the rod domain.
The phylogenic analysis implies that the multiple -actinin isoforms found in modern vertebrates arose after the divergence of the vertebrate and urochordate lineages. The determined phylogeny displayed an ((A, B) (C, D)) topology, implying that the calcium-insensitive -actinin isoforms 2 and 3 evolved together as did the calcium-sensitive isoforms 1 and 4.
It is believed that an -actinin–like precursor has given rise to members of the spectrin superfamily (i.e., -actinins, spectrin, and dystrophin) by gene duplications and gene rearrangements (Pascual, Castresana, and Saraste 1997; Thomas et al. 1997; Viel 1999; Baines 2003). Therefore, a better understanding of the evolution and structure of -actinins may provide a better insight into the evolution of this superfamily.
Materials and Methods
A Ciona -Zap cDNA library (generously made available by Dr. Nori Satoh) made from tail-bud embryos was screened using the following -actinin–specific oligonucleotides: N-term forward: 5'-CAGGAGGAGGAGTGGGACCGCG-3 and N-term reverse: 5'-CCCATACTAGACCTGGTAGTAGG-3' (corresponding to actin-binding domain of human -actinin); forward I: 5'-GGTGGCTTGACAAGGAAACTAG-3' and forward II: 5'-GATGGTTTGGCTTTCTGTGCC C-3'; and reverse I: 5'-CTCCTGGTTCAAACCAAGAACCTTC-3' and reverse II: 5'-GCCTTGACT TCTTGGAAGGTGC-3' (all specific to the downstream region of Ciona -actinin). M13-specific primers (M13 forward: 5'-CGTTGTAAAACGACGGCCAGTG-3' and M13 reverse: 5'-GCATT-3' AGTACCAGTATCGACAAAGGAC-3') were also used. PCR Master mix (Promega) was used for the PCR reactions. Obtained PCR products were purified by QIAquick PCR purification kit (Qiagen) and ligated into pGEM-T vector (Promega). After the ligation, TG1 E. coli cells were transformed by heat shock and cultured on agar plates overnight at 37°C. Isolated clones were sequenced using the ABI PRISM BigDye terminator cycle sequencing kit (Applied Biosystems).
Phylogenetic Analysis
Available nucleotide and amino acid sequences of -actinins were retrieved from NCBI, TIGR, Swiss Protein and Sanger Institute. Retrieved sequences are summarized in table 1.
Table 1 Accession Number, Protein Name, and Phylum of -Actinins Used in the Phylogenetic Analysis.
Sequences were aligned in ClustalX (Thompson et al. 1997), using default parameters, and refined manually. Phylogenetic analysis was conducted using maximum-parsimony, neighbor-joining, and Bayesian inference; they all resulted in very similar results, both in topology and in significance. The analyses were done using the PHYLIP (Felsenstein 1989), MEGA2 (Kumar et al. 2001), and MrBayes software packages (Huelsenbeck and Ronquist 2001).
One-thousand bootstrapped data sets were created with the program SEQBOOT, and the phylogeny estimate for each data was calculated using PROTPARS, selecting 1,000 data sets. The input order of the sequences was randomized with a jumble number of 10. The majority-rule consensus tree was created by CONSENSUS.
Results
Ciona intestinalis -Actinin
A Ciona cDNA library was screened by PCR for the presence of -actinin. The primers used for this purpose were based on the 5'-end of the human -actinin nucleotide sequence, corresponding to the highly conserved N-terminal domain. Subcloning and subsequent sequencing of obtained PCR products identified several overlapping DNA fragments. The translated amino acid sequence was highly similar to -actinin of vertebrates.
During the course of this work, part of the Ciona genome was published (Dehal et al. 2002; Satou et al. 2002). Alignment of our sequence with the TIGR Ciona intestinalis Gene Index, pulled out TC14429, identified as a partial fragment of the -actinin (97%) similar to Drosophila -actinin. We observed that the sequence we obtained was extended upstream of the probable start codon, completing the sequence. From this sequence new specific primers for the Ciona -actinin were designed, and a longer sequence was obtained that overlapped completely with the sequence in the TIGR database.
As our PCR results imply, Ciona appears to have only a single gene for -actinin. Previous comparative studies of gene families in Ciona have show that most genes that are present in multiple copies in vertebrates only have a single representative in Ciona (Dehal et al. 2002).
-Actinin from Other Organisms
Available databases (e.g., NCBI, SwissProt, TIGR, and Sanger Institute) were searched for sequences similar to Ciona -actinin, as well as for annotated -actinin sequences. We retrieved sequences of -actinin from 17 different species, mostly from the animal kingdom. Several of the vertebrates have more than one isoform. In humans and mouse, four different isoforms have been identified, whereas in rat and chicken, only three distinct isoforms have been identified so far. In total, we retrieved 26 -actinin sequences (table 1). These sequences were first aligned using ClustalX, with default settings, and then refined manually.
It is obvious that throughout the vertebrates, all -actinins are very similar; 93% (or 837 of 892) of the amino acid residues are identical in human and frog -actinin. The isoforms of a single species are also highly similar; of the approximately 890 amino acid residues in the four isoforms of human -actinins 73% or more are identical (fig. 1).
FIG. 1. Pairwise sequence identity of -actinin isoforms
Although the unicellular -actinins are less similar, the degree of identity is still very high, particular in the N-terminus and C-terminus. The largest sequence differences are found in the rod domain. In all vertebrate sequences, it is possible to locate four spectrin repeats, whereas the rod domains of Schizosaccharomyces pombe and Neurospora crassa contain two spectrin repeats, and those of Entamoeba histolytica and Encephalitozoo cuniculi have one repeat.
Schizosaccharomyces pombe and Neurospora crassa -Actinins
S. pombe and N. crassa are both ascomycetes belonging to the fungus kingdom (Keeling, Luker, and Palmer 2000). As expected, the -actinins of these two organisms display a very high degree of identity. When compared with other -actinins, it is obvious that the calponin homology (actin-binding) domain in the N-terminus and the calcium-binding domain in the C-terminus have been preserved during evolution. However, the rod domain in both S. pombe and N. crassa -actinins differs. This part is much shorter than in other -actinins and comprises only two spectrin repeats. These repeats are similar to the first repeat (27% identity) and fourth repeat (21% identity), respectively, of Drosophila -actinin (Wu, Bahler, and Pringle 2001).
Entamoeba histolytica -actinin–like Protein
Entamoeba histolytica is considered one of the most primitive protozoa. Although it is a eukaryotic organism, it lacks mitochondria and several other characteristic organelles. E. histolytica is parasitic and infects predominantly humans and other primates (Nickel et al. 2000). A protein that cross-reacts with antibodies to -actinin has been identified (Bailey et al. 1992). From the alignment of this -actinin–like protein, the typical domain structure is not obvious. When the sequence (Nickel et al. 2000) was submitted either to InterPro scan or SMART, an actin-binding motif and two calponin homology domains in the N-terminus as well as two, or possibly three EF-hand motifs in the C-terminus were returned. Using the ProteinPredict server, the actin-binding domain could be modeled on the coordinates of the actin-binding domain of human dystrophin (1dxx) and the calcium-binding domain of porcine calponin (1alv).
The rod domain of this protein is much shorter than that of the prototypic -actinin. It comprises only approximately 125 amino acid residues, which should correspond to a single spectrin repeat. This repeat was found to be most closely related to the first repeat of chicken -actinin.
Encephalitozoo cuniculi -Actinin
Encephalitozoo cuniculi is an obligate intracellular parasite, belonging to the microsporidia (Weiss 2001). Similar to E. histolytica, microsporidia are characterized by the lack of normal mitochondria and peroxisomes (Roger and Silberman 2002; Williams et al. 2002). The taxonomy of this group has been reclassified several times and is still uncertain (Germot, Philippe, and Le Guyader 1997; Peyretaillade et al. 1998; Keeling, Luker, and Palmer 2000; Keeling and Fast 2002; Keeling 2003). Based on phylogenetic analysis of small subunit ribosomal RNA as well as the lack of mitochondria, it has been suggested that microsporidia are one of the earliest eukaryote lineages (Leipe et al. 1993). More recent analysis indicate that microsporidia are related to fungi and not to early diverging eukaryotes (Hirt et al. 1999; Keeling, Luker, and Palmer 2000; Keeling 2003).
Similar to the fungi isoforms, the -actinin of E. cuniculi is short; it is only 537 amino acid residues long. The similarity to the prototypic -actinin is low (fig. 1). When submitted to the InterPro Scan and SMART, an actin-binding motif and two calponin homology domains in the N-terminal domain were identified. Further analysis of the primary structure indicated the highest similarity with sequences from fungi and amoebae (fig. 1).
Trichomonas vaginalis -Actinin
Trichomonas vaginalis is a sexually transmitted human parasite. When it adheres to the host cell, it undergoes a transformation from a flagellate to an amoeboid form. Previous analysis of the -actinin of T. vaginalis has suggested the presence of five repeats in the rod domain in addition to the common motifs in both termini (Addis et al. 1998; Bricheux et al. 1998). Interesting to note is that only the first of these repeats shows some similarity with other -actinins, and then only with the first repeat of other -actinins. It was suggested that the other four repeats have evolved because of intragenic duplication that has not occurred in other sequences (Bricheux et al. 1998).
Phylogenetic Analysis
The aligned sequences were analyzed using the PHYLIP, MEGA2, and MrBayes packages. The program SEQBOOT was used to create 1,000 bootstrapped data set and PROTPARS was used to calculate the phylogeny estimate. The maximum-parsimony tree obtained is shown in figure 2. Phylogenetic analysis based on neighbor-joining and Bayesian inference resulted in trees with very similar topologies as well as significance.
FIG. 2. Phylogenetic tree of -actinins. The most-parsimonious tree shows the relation between -actinins of different organism, as well as of different isoforms of the same organism. Bootstrap values in percentage were calculated from 1,000 data sets
There are three main branches in the tree; two vertebrate branches and one invertebrate and unicellular organisms branch. One of the vertebrate branches includes the calcium-insensitive isoforms 2 and 3, and the other branch includes the calcium-sensitive isoforms 1 and 4, thus giving rise to a phylogenetic tree with an ((A, B) (C, D)) topology. That -actinins of each isoforms are grouped together implies a higher degree of similarity between a particular isoform of all organisms than between the isoforms of a single organism. Most branches in the maximum-parsimony consensus tree are supported by bootstrap values of 90% or higher.
The phylogeny was also determined on a subset of the -actinin sequences. Analyzing the amino acid sequences of the actin-binding domain of the vertebrate -actinins resulted in very similar tree topologies. However, when the nucleotide sequences were used for the analysis instead, an (A, (B, (C, D))) topology was obtained.
In the invertebrate branch Ciona -actinin appears to be closest related to -actinins of the arthropod branch (Drosophila and Anopheles) and the nematode C. elegans. The placing of unicellular -actinins at the end of this branch indicates that these have branched off early in evolution.
Analysis of the Spectrin Repeats
We also analyzed the phylogenies of the spectrin repeats. All methods tested placed the single spectrin repeat of E. histolytica -actinins closest to the first spectrin repeat (SR1), as shown in figure 3. The two repeats of S. pombe and N. crassa are most similar to SR1 and SR4, respectively. This suggests a possible evolutionary pathway (fig. 4). In the primordial ancestor to present-day -actinins, the rod domain contained only a single spectrin repeat that became the first repeat (SR1) in modern -actinins. A first intragenic duplication gave rise to the second repeat, which became the fourth repeat (SR4) in modern -actinins. A subsequent second intragenic duplication added two more repeats (SR2 and SR3).
FIG. 3. Phylogenetic analysis of spectrin repeats in chicken, yeast, and amoeba. In the most-parsimonious tree, the single repeat of Enthamoeba histolytica (entamoeba) appears closest to the first repeat of chicken (SR1) -actinin. Similarly, the first repeat of the Schizosaccharomyces pombe isoform (pombe SR1) appears most related to the first repeat in the chicken isoform (SR1), whereas the second repeat of yeast -actinin (pombe SR2) seems to be closest to the fourth chicken repeat (SR4). The bootstrap values in percentage were calculated from 1,000 data sets
FIG. 4. Schematic structure of ancestral and modern -actinins. All -actinins have preserved the two calponin homology domains and, in most cases, the calcium-binding domain, whereas the rod domain differs significantly between different isoforms. Basal organisms, such as Entamoeba histolytica and Schizosaccharomyce pombe with one and two spectrin repeats, respectively, in the rod domain indicate that the two repeats in S. pombe and the four repeats present in other organisms are the result of two consecutive intragenic duplications from the single spectrin repeat of a primordial -actinin, like the one present in E. histolytica
Discussion
The existence of a notochord in the larvae stage implies that the urochordate Ciona share a common ancestor with vertebrates before the vertebrate radiation (Dehal et al. 2002; Gee 2002; Pennisi 2002). Since Ciona appears to have only a single -actinin gene in contrast to higher vertebrates such as human, mouse, and chicken, it seems likely that the appearance of the different -actinin isoforms arose after the vertebrate radiation.
This theory is substantiated by the topology of the phylogenetic tree, which indicated one invertebrate and two major vertebrate branches. One of the vertebrate branches encompasses the calcium-insensitive -actinin 2 and -actinin 3, and the other branch encompasses the calcium-sensitive isoforms 1 and 4. This ((A, B) (C, D)) tree topology lends support to the suggestion that two rounds of genome duplication occurred early in vertebrate evolution (the 2R hypothesis) (Hughes 1999; Durand 2003). However, the observed phylogeny can also be explained by independent duplication of individual genes or gene segments.
A phylogenetic analysis using the nucleotide sequences corresponding to the actin-binding domain of vertebrate -actinins resulted in a different tree topology (Dixson, Forstner, and Garcia 2003). In this case, the tree displayed a (A, (B, (C, D))) topology, suggesting that the calcium-insensitive -actinin 2 evolved initially, followed first by the calcium-insensitive -actinin 3 and then by the calcium-sensitive -actinins 1 and 4. When we repeated the phylogenetic analysis on this subset of sequence data, using the amino acid sequence of the actin-binding domain, we obtained tree topologies very similar to those obtained for the full data set. However, when we repeated the analysis using the nucleotide sequences, we obtained an (A, (B, (C, D))) tree topology, similar to that reported by Dixson, Forstner, and Garcia (2003). Thus, depending on whether amino acid or nucleotide sequences are used in the phylogenetic analysis, different tree topologies may be obtained.
It has been suggested that phylogenetic analysis based on amino acid sequences are more reliable than analysis based on nucleotide sequences because of compositional bias in the DNA sequences (Foster and Hickey 1999). This, together with the high bootstrap values, gives substantial support to our phylogenetic analysis.
Also frog (X. laevis) and fish (D. rerio and F. rubripes) appear to have a single -actinin isoform. The frog and puffer fish -actinins are related to isoform 1, whereas the -actinin of zebrafish is related to -actinin 3. The appearance of only a single isoforms in each of these organisms does not exclude the possibility of other isoforms, as their genomes have not yet been completely sequenced.
All tested phylogeny models placed the isoform of E. histolytica as the earliest diverging -actinin, followed by the -actinin of E. cuniculi. Together with the isoform of T. vaginalis, these two -actinins are also those that are least similar to the consensus sequence. Despite the low similarity, these proteins display certain of the typical hallmarks of -actinins and can be regarded as -actinins or at least as -actinin–like proteins.
The fungi S. pombe and N. crassa also branched off early in evolution. The fungi -actinins are more similar to the present day -actinins, despite the short rod domain with its two spectrin repeats.
When aligning the sequence of E. cuniculi -actinin with -actinins of E. histolytica and S. pombe, it is apparent that the sequence is similar not only to the S. pombe isoform but also to the E. histolytica -actinin, again placing microsporidia between the fungus and protozoa branches.
In all bacterial and plant genomes available to date, we have not been able to identify any genes for a -actinin or a -actinin–like protein. This implies that the primordial -actinin appeared in a primitive unicellular organism belonging to the protozoan kingdom.
-Actinin is composed of three major domains: an N-terminal actin-binding domain, consisting of two calponin homology domains; a rod domain of spectrin repeats, and a C-terminal calcium-binding domain. These domains are highly conserved throughout the animal kingdom. However, in basal eukaryotes and, in particular, in unicellular organisms, this domain structure is less well preserved. The evolutionary best-preserved domain is the N-terminal domain with its two calponin homology domains, which is present in all -actinins and -actinin–like proteins. This is not surprising, as the major function of -actinin is believed to be binding actin filaments (Blanchard, Ohanian, and Critchley 1989). Although it was not possible to identify a calcium-binding motif in all sequences, it is apparent that this domain is also well conserved among the different -actinins. It should be noticed that not all isoforms of -actinin are sensitive to calcium ions because of structural changes in this region during evolution (Dixson, Forstner, and Garcia 2003).
The most variable part is the rod domain, which is composed of one, two, or four spectrin repeats. In -actinins able to form antiparallel dimers, the function of the rod domain is mainly to separate the actin-binding domains of each molecule (Ylanne et al. 2001). Therefore, the rod part could be built from any amino acid sequence that can form a coiled-coil or rodlike structure. The -actinin of the protozoa Trichomonas vaginalis seems to support this conclusion, as only one spectrin repeat can be identified in the rod domain in addition to a long coiled-coil region. Also, the rod domain of E. histolytica -actinin contains only a single repeat unit, but in this case, there is no additional coiled-coil region.
The rod domain of the isoforms of E. histolytica and E. cuniculi are much shorter, and only a single spectrin repeat could be identified. Although the rod domain of T. vaginalis -actinin is longer, also in this case, only a single repeat could be identified.
The rod domain of -actinins of all vertebrates, as well as of invertebrates, has four spectrin repeats. This suggests a likely scenario for the evolution of the rod domain. In the primordial -actinin, the rod domain was short, spanning a single spectrin repeat. An intragenic duplication added a second repeat, followed by yet another intragenic duplication. As the single repeat in E. histolytica appears most related to the first repeat (SR1), it is likely that this repeat also was the first to evolve. Likewise, as the two repeats of fungi are closest to repeat 1 (SR1) and 4 (SR4), this suggests that repeat 4 aroused through intragenic duplication. The second intragenic duplication added repeats 2 (SR2) and 3 (SR3). Thus, our data do not support the conclusion that the ancestor of -actinin should possess four spectrin repeats (Viel 1999).
It has been demonstrated that SR2 and SR3 are essential for dimerization of -actinin (Djinovic-Carugo et al. 1999). In the dimer, SR2 and SR3 pair with the SR3 and SR2 of the other antiparallel molecule. Therefore, it is possible that these primitive -actinins are unable to form dimers, which would exclude any function involving actin cross-linking. The function of these short and atypical -actinins is still unknown.
Supplementary Material
The final alignment of all retrieved -actinin sequences is available online at www.mbe.oupjournals.org.
Acknowledgements
This work was supported by grants from EU HPRN-CT-2000–00096, ke Wibergs Stiftelse, Magn. Bergvalls Stiftelse and O.E. och Edla Johansson Stiftelse.
The TIGR Ciona intestinalis Gene Index updated February 2, 2004, now contains clone TC35693, which is identical to the Ciona -actinin sequence determined in this work.
Literature Cited
Addis, M. F., P. Rappelli, G. Delogu, F. Carta, P. Cappuccinelli, and P. L. Fiori. 1998. Cloning and molecular characterization of a cDNA clone coding for Trichomonas vaginalis alpha-actinin and intracellular localization of the protein. Infect. Immun. 66:4924-4931.
Bailey, G. B., P. S. Shen, M. J. Beanan, and N. E. McCoomer. 1992. Actin associated proteins of Entamoeba histolytica. Arch. Med. Res. 23:129-132.
Baines, A. J. 2003. Comprehensive analysis of all triple helical repeats in beta-spectrins reveals patterns of selective evolutionary conservation. Cell. Mol. Biol. Lett. 8:195-214.
Barstead, R. J., L. Kleiman, and R. H. Waterston. 1991. Cloning, sequencing and mapping of an alpha-actinin gene from the nematode Caenorhabditis Elegans. Cell Motil. Cytoskelet. 20:69-78.
Beggs, A. H., T. J. Byers, J. H. M. Knoll, F. M. Boyce, G. A. P. Bruns, and L. M. Kunkel. 1992. Cloning and characterization of two human skeletal muscle alpha-actinin genes located on chromosomes 1 and 11. J. Biol. Chem. 267:9281-9288.
Blanchard, A., V. Ohanian, and D. Critchley. 1989. The structure and function of alpha-actinin. J. Muscle Res. Cell Motil. 10:280-289.
Bricheux, G., G. Coffe, N. Pradel, and G. Brugerolle. 1998. Evidence for an uncommon -actinin protein in Trichomonas vaginalis. Mol. Biochem. Parasitol. 95:241-249.
Dehal, P., Y. Satou, and R. K. Campbell, et al. (87 co-authors). 2002. The draft genome of Ciona intestinalis: insights into chordate and vertebrate origins. Science. 298:2157-2167.
Dixson, J. D., M. J. Forstner, and D. M. Garcia. 2003. The -actinin gene family: A revised classification. J. Mol. Evol. 56:1-10.
Djinovic-Carugo, K., M. Gautel, J. Ylanne, and P. Young. 2002. The spectrin repeat: a structural platform for cytoskeletal protein assemblies. FEBS Lett. 513:119-123.
Djinovic-Carugo, K., P. Young, M. Gautel, and M. Saraste. 1999. Structure of the -actinin rod: molecular basis for cross-linking of actin filaments. Cell. 98:537-546.
Dubreuil, R. R. 1991. Structure and evolution of the actin crosslinking proteins. Bioessays. 13:219-226.
Durand, D. 2003. Vertebrate evolution: doubling and shuffling with a full deck. Trends Genet. 19:2-5.
Felsenstein, J. 1989. PHYLIP (phylogeny inference package) Version 3.2. Caldistics. 5:164-166.
Foster, P. G., and D. A. Hickey. 1999. Compositional bias may affect both DNA-based and protein-based phylogenetic reconstructions. J. Mol. Evol. 48:284-290.
Fyrberg, E., M. Kelly, E. Ball, C. Fyrberg, and M. C. Reedy. 1990. Molecular genetics of Drosophila alpha-actinin: mutant alleles disrupt Z disc integrity and muscle insertions. J. Cell Biol. 110:1999-2011.
Gee, H. 2002. Genomics: return of a little squirt. Nature. 420:755-756.
Germot, A., H. Philippe, and H. Le Guyader. 1997. Evidence for loss of mitochondria in Microsporidia from a mitochondrial-type HSP70 in Nosema locustae. Mol. Biochem. Parasitol. 87:159-168.
Hirt, R. P., J. M. Logsdon, Jr., B. Healy, M. W. Dorey, W. F. Doolittle, and T. M. Embley. 1999. Microsporidia are related to Fungi: evidence from the largest subunit of RNA polymerase II and other proteins. Proc. Natl. Acad. Sci. USA 96:580-585.
Huelsenbeck, J. P., and F. Ronquist. 2001. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17:754-755.
Hughes, A. L. 1999. Phylogenies of developmentally important proteins do not support the hypothesis of two rounds of genome duplication early in vertebrate history. J. Mol. Evol. 48:565-576.
Janssen, K. P., L. Eichinger, P. A. Janmey, A. A. Noegel, M. Schliwa, W. Witke, and M. Schleicher. 1996. Viscoelastic properties of F-actin solutions in the presence of normal and mutated actin-binding proteins. Arch. Biochem. Biophys. 325:183-189.
Keeling, P. J. 2003. Congruent evidence from alpha-tubulin and beta-tubulin gene phylogenies for a zygomycete origin of microsporidia. Fungal. Genet. Biol. 38:298-309.
Keeling, P. J., and N. M. Fast. 2002. Microsporidia: biology and evolution of highly reduced intracellular parasites. Annu. Rev. Microbiol. 56:93-116.
Keeling, P. J., M. A. Luker, and J. D. Palmer. 2000. Evidence from beta-tubulin phylogeny that microsporidia evolved from within the fungi. Mol. Biol. Evol. 17:23-31.
Kumar, S., K. Tamura, I. B. Jakobsen, and M. Nei. 2001. MEGA2: molecular evolutionary genetics analysis software. Bioinformatics. 17:1244-1245.
Leipe, D. D., J. H. Gunderson, T. A. Nerad, and M. L. Sogin. 1993. Small subunit ribosomal RNA+ of Hexamita inflata and the quest for the first branch in the eukaryotic tree. Mol. Biochem. Parasitol. 59:41-48.
Mills, M., N. Yang, R. Weinberger, D. L. Vander Woude, A. H. Beggs, S. Easteal, and K. North. 2001. Differential expression of the actin-binding proteins, alpha-actinin-2 and -3, in different species: implications for the evolution of functional redundancy. Hum. Mol. Genet. 10:1335-1346.
Nickel, R., T. Jacobs, B. Urban, H. Scholze, H. Bruhn, and M. Leippe. 2000. Two novel calcium-binding proteins from cytoplasmic granules of the protozoan parasite Entamoeba histolytica. FEBS Lett. 486:112-116.
Norwood, F. L. M., A. J. Sutherland-Smith, N. H. Keep, and J. Kendrick-Jones. 2000. The structure of the N-terminal actin-binding domain of human dystrophin and how mutations in this domain may cause Duchenne or Becker muscular dystrophy. Struct. Fold Des. 8:481-491.
Pascual, J., J. Castresana, and M. Saraste. 1997. Evolution of the spectrin repeat. BioEssays. 19:811-817.
Pennisi, E. 2002. Comparative genomics: tunicate genome shows a little backbone. Science. 298:2111-2112.
Peyretaillade, E., V. Broussolle, P. Peyret, G. Metenier, M. Gouy, and C.P. Vivares. 1998. Microsporidia, amitochondrial protists, possess a 70-kDa heat shock protein gene of mitochondrial evolutionary origin. Mol. Biol. Evol. 15:683-689.
Roger, A. J., and J. D. Silberman. 2002. Cell evolution: mitochondria in hiding. Nature. 418:827-829.
Satou, Y., L. Yamada, and Y. Mochizuki, et al. (14 co-authors). 2002. A cDNA resource from the basal chordate Ciona intestinalis. Genesis. 33:153-154.
Tang, J., D. W. Taylor, and K. A. Taylor. 2001. The three-dimensional structure of alpha-actinin obtained by cryoelectron microscopy suggests a model for Ca2+-dependent actin binding. J. Mol. Biol. 310:845-858.
Thomas, G. H., E. C. Newbern, C. C. Korte, M. A. Bales, S. V. Muse, A. G. Clark, and D. P. Kiehart. 1997. Intragenic duplication and divergence in the spectrin superfamily of proteins. Mol. Biol. Evol. 14:1285-1295.
Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25:4876-4882.
Viel, A. 1999. -Actinin and spectrin structures: an unfolding family story. FEBS Lett. 460:391-394.
Weiss, L. M. 2001. Microsporidia: emerging pathogenic protists. Acta Trop. 78:89-102.
Williams, B. A., R. P. Hirt, J. M. Lucocq, and T. M. Embley. 2002. A mitochondrial remnant in the microsporidian Trachipleistophora hominis. Nature. 418:865-869.
Witke, W., A. Hofmann, B. Koeppel, M. Schleicher, and A. A. Noegel. 1993. The Ca2+-binding domains in non-muscle type -actinin: biochemical and genetic analysis. J. Cell Biol. 121:599-606.
Wu, J. Q., J. Bahler, and J. R. Pringle. 2001. Roles of a fimbrin and an -actinin-like protein in fission yeast cell polarization and cytokinesis. Mol. Biol. Cell. 12:1061-1077.
Ylanne, J., K. Scheffzek, P. Young, and M. Saraste. 2001. Crystal structure of the alpha-actinin rod reveals an extensive torsional twist. Structure (Camb). 9:597-604.(Ana Virel and Lars Backma)
E-mail: lars.backman@chem.umu.se.
Abstract
The N-terminal actin-binding domain of -actinin is connected to the C-terminal EF-hands by a rod domain. Because of its ability to form dimers, -actinin can cross-link actin filaments in muscle cells as well as in nonmuscle cells. In the prototypic -actinins, the rod domain contains four triple helical bundles, or so-called spectrin repeats. We have found some atypical -actinins in early diverging organisms, such as protozoa and yeast, where the rod domain contains one and two spectrin repeats, respectively. This implies that the four repeats present in modern -actinins arose after two consecutive intragenic duplications from an -actinin with a single repeat. Further, the evolutionary gene tree of -actinins shows that the appearance of four distinct -actinin isoforms may have occurred after the vertebrate-invertebrate split. The topology of the tree lends support to the hypothesis that two rounds (2R) of genome duplication occurred early in the vertebrate radiation. The phylogeny also considers these atypical isoforms as the most basal to -actinins of vertebrates and other eukaryotes. The analysis also positioned -actinin of the fungi Encephalitozoo cuniculi close to the protozoa, supporting the suggestion that microsporidia are early eukaryotes. Because -actinin is considered the basal member of the spectrin family, our studies will improve the understanding of the origin and evolution of this superfamily.
Key Words: -actinin ? phylogeny ? evolution ? spectrin superfamily ? spectrin repeat
Introduction
-Actinin is a ubiquitous actin cross-linker belonging to the spectrin superfamily (Blanchard, Ohanian, and Critchley 1989; Dubreuil 1991). It has been identified in most eukaryotic organisms, from human (Beggs et al. 1992; Mills et al. 2001) and mouse (Mills et al. 2001) to fly (Drosophila melanogaster) and worm (Caenorhabditis elegans) (Fyrberg et al. 1990; Barstead, Kleiman, and Waterston 1991). Although fission yeast (Schizosaccharomyces pombe) appears to express an -actinin (or at least an -actinin–like protein) baker's yeast (Saccharomyces cerevisiae) does not (Wu, Bahler, and Pringle 2001).
The structure of the -actinin has been determined in a great detail. Depending on the source, this elongated protein has a molecular mass of 93 to 103 kDa (Blanchard, Ohanian, and Critchley 1989). The actin-binding site is located at the N-terminus, which comprises two calponin homology domains, where only the most N-terminal one binds actin (Norwood et al. 2000). At the C-terminus, there are two EF-hands, where the second EF-hand is crucial for calcium binding (Witke et al. 1993; Janssen et al. 1996). These two domains are connected by the rod domain. This domain is formed by triple-helical repeats or so-called spectrin repeats, where each repeat contains approximately 106 amino acid residues (Djinovic-Carugo et al. 1999). The rod domain of -actinin is important for dimerization (Djinovic-Carugo et al. 1999; Ylanne et al. 2001). There are usually four spectrin repeats in the rod domain, but -actinin from some organisms appears to have only one or two repeats.
Because the rod domain determines the distance between cross-linked actin filaments and serves as an interaction site for receptors and adaptor proteins, the number of repeats has implications for actin bundling as well as for cytoskeleton organization (Djinovic-Carugo et al. 1999; Ylanne et al. 2001; Djinovic-Carugo et al. 2002).
-Actinins are classified into four different main isoforms (Dixson, Forstner, and Garcia 2003). Humans and probably other vertebrates express all four isoforms, whereas invertebrates and protista seem to express only one isoform. The muscle isoforms, -actinin 2 and of -actinin 3 are localized to the Z-disc of the sarcomeres (Blanchard, Ohanian, and Critchley 1989; Mills et al. 2001). These two isoforms are insensitive to calcium because their EF-hands are nonfunctional (Beggs et al. 1992). In contrast, the nonmuscle isoforms, -actinin 1 and of -actinin 4 are sensitive to calcium (Tang, Taylor, and Taylor 2001). -Actinin 1 and of -actinin 4 can be found at the leading edge of motile cells, at cell adhesion sites and focal contacts, and along actin stress fibers in migrating cells (Barstead, Kleiman, and Waterston 1991).
To better understand the structural and evolutionary relationships between the -actinin isoforms, we have characterized -actinin from the urochordate Ciona intestinalis, as well as several other primitive eukaryotes. Our results demonstrate that the general structural pattern of -actinin is not applicable to the most primitive eukaryotes, as these isoforms are shorter because of fewer spectrin repeats in the rod domain.
The phylogenic analysis implies that the multiple -actinin isoforms found in modern vertebrates arose after the divergence of the vertebrate and urochordate lineages. The determined phylogeny displayed an ((A, B) (C, D)) topology, implying that the calcium-insensitive -actinin isoforms 2 and 3 evolved together as did the calcium-sensitive isoforms 1 and 4.
It is believed that an -actinin–like precursor has given rise to members of the spectrin superfamily (i.e., -actinins, spectrin, and dystrophin) by gene duplications and gene rearrangements (Pascual, Castresana, and Saraste 1997; Thomas et al. 1997; Viel 1999; Baines 2003). Therefore, a better understanding of the evolution and structure of -actinins may provide a better insight into the evolution of this superfamily.
Materials and Methods
A Ciona -Zap cDNA library (generously made available by Dr. Nori Satoh) made from tail-bud embryos was screened using the following -actinin–specific oligonucleotides: N-term forward: 5'-CAGGAGGAGGAGTGGGACCGCG-3 and N-term reverse: 5'-CCCATACTAGACCTGGTAGTAGG-3' (corresponding to actin-binding domain of human -actinin); forward I: 5'-GGTGGCTTGACAAGGAAACTAG-3' and forward II: 5'-GATGGTTTGGCTTTCTGTGCC C-3'; and reverse I: 5'-CTCCTGGTTCAAACCAAGAACCTTC-3' and reverse II: 5'-GCCTTGACT TCTTGGAAGGTGC-3' (all specific to the downstream region of Ciona -actinin). M13-specific primers (M13 forward: 5'-CGTTGTAAAACGACGGCCAGTG-3' and M13 reverse: 5'-GCATT-3' AGTACCAGTATCGACAAAGGAC-3') were also used. PCR Master mix (Promega) was used for the PCR reactions. Obtained PCR products were purified by QIAquick PCR purification kit (Qiagen) and ligated into pGEM-T vector (Promega). After the ligation, TG1 E. coli cells were transformed by heat shock and cultured on agar plates overnight at 37°C. Isolated clones were sequenced using the ABI PRISM BigDye terminator cycle sequencing kit (Applied Biosystems).
Phylogenetic Analysis
Available nucleotide and amino acid sequences of -actinins were retrieved from NCBI, TIGR, Swiss Protein and Sanger Institute. Retrieved sequences are summarized in table 1.
Table 1 Accession Number, Protein Name, and Phylum of -Actinins Used in the Phylogenetic Analysis.
Sequences were aligned in ClustalX (Thompson et al. 1997), using default parameters, and refined manually. Phylogenetic analysis was conducted using maximum-parsimony, neighbor-joining, and Bayesian inference; they all resulted in very similar results, both in topology and in significance. The analyses were done using the PHYLIP (Felsenstein 1989), MEGA2 (Kumar et al. 2001), and MrBayes software packages (Huelsenbeck and Ronquist 2001).
One-thousand bootstrapped data sets were created with the program SEQBOOT, and the phylogeny estimate for each data was calculated using PROTPARS, selecting 1,000 data sets. The input order of the sequences was randomized with a jumble number of 10. The majority-rule consensus tree was created by CONSENSUS.
Results
Ciona intestinalis -Actinin
A Ciona cDNA library was screened by PCR for the presence of -actinin. The primers used for this purpose were based on the 5'-end of the human -actinin nucleotide sequence, corresponding to the highly conserved N-terminal domain. Subcloning and subsequent sequencing of obtained PCR products identified several overlapping DNA fragments. The translated amino acid sequence was highly similar to -actinin of vertebrates.
During the course of this work, part of the Ciona genome was published (Dehal et al. 2002; Satou et al. 2002). Alignment of our sequence with the TIGR Ciona intestinalis Gene Index, pulled out TC14429, identified as a partial fragment of the -actinin (97%) similar to Drosophila -actinin. We observed that the sequence we obtained was extended upstream of the probable start codon, completing the sequence. From this sequence new specific primers for the Ciona -actinin were designed, and a longer sequence was obtained that overlapped completely with the sequence in the TIGR database.
As our PCR results imply, Ciona appears to have only a single gene for -actinin. Previous comparative studies of gene families in Ciona have show that most genes that are present in multiple copies in vertebrates only have a single representative in Ciona (Dehal et al. 2002).
-Actinin from Other Organisms
Available databases (e.g., NCBI, SwissProt, TIGR, and Sanger Institute) were searched for sequences similar to Ciona -actinin, as well as for annotated -actinin sequences. We retrieved sequences of -actinin from 17 different species, mostly from the animal kingdom. Several of the vertebrates have more than one isoform. In humans and mouse, four different isoforms have been identified, whereas in rat and chicken, only three distinct isoforms have been identified so far. In total, we retrieved 26 -actinin sequences (table 1). These sequences were first aligned using ClustalX, with default settings, and then refined manually.
It is obvious that throughout the vertebrates, all -actinins are very similar; 93% (or 837 of 892) of the amino acid residues are identical in human and frog -actinin. The isoforms of a single species are also highly similar; of the approximately 890 amino acid residues in the four isoforms of human -actinins 73% or more are identical (fig. 1).
FIG. 1. Pairwise sequence identity of -actinin isoforms
Although the unicellular -actinins are less similar, the degree of identity is still very high, particular in the N-terminus and C-terminus. The largest sequence differences are found in the rod domain. In all vertebrate sequences, it is possible to locate four spectrin repeats, whereas the rod domains of Schizosaccharomyces pombe and Neurospora crassa contain two spectrin repeats, and those of Entamoeba histolytica and Encephalitozoo cuniculi have one repeat.
Schizosaccharomyces pombe and Neurospora crassa -Actinins
S. pombe and N. crassa are both ascomycetes belonging to the fungus kingdom (Keeling, Luker, and Palmer 2000). As expected, the -actinins of these two organisms display a very high degree of identity. When compared with other -actinins, it is obvious that the calponin homology (actin-binding) domain in the N-terminus and the calcium-binding domain in the C-terminus have been preserved during evolution. However, the rod domain in both S. pombe and N. crassa -actinins differs. This part is much shorter than in other -actinins and comprises only two spectrin repeats. These repeats are similar to the first repeat (27% identity) and fourth repeat (21% identity), respectively, of Drosophila -actinin (Wu, Bahler, and Pringle 2001).
Entamoeba histolytica -actinin–like Protein
Entamoeba histolytica is considered one of the most primitive protozoa. Although it is a eukaryotic organism, it lacks mitochondria and several other characteristic organelles. E. histolytica is parasitic and infects predominantly humans and other primates (Nickel et al. 2000). A protein that cross-reacts with antibodies to -actinin has been identified (Bailey et al. 1992). From the alignment of this -actinin–like protein, the typical domain structure is not obvious. When the sequence (Nickel et al. 2000) was submitted either to InterPro scan or SMART, an actin-binding motif and two calponin homology domains in the N-terminus as well as two, or possibly three EF-hand motifs in the C-terminus were returned. Using the ProteinPredict server, the actin-binding domain could be modeled on the coordinates of the actin-binding domain of human dystrophin (1dxx) and the calcium-binding domain of porcine calponin (1alv).
The rod domain of this protein is much shorter than that of the prototypic -actinin. It comprises only approximately 125 amino acid residues, which should correspond to a single spectrin repeat. This repeat was found to be most closely related to the first repeat of chicken -actinin.
Encephalitozoo cuniculi -Actinin
Encephalitozoo cuniculi is an obligate intracellular parasite, belonging to the microsporidia (Weiss 2001). Similar to E. histolytica, microsporidia are characterized by the lack of normal mitochondria and peroxisomes (Roger and Silberman 2002; Williams et al. 2002). The taxonomy of this group has been reclassified several times and is still uncertain (Germot, Philippe, and Le Guyader 1997; Peyretaillade et al. 1998; Keeling, Luker, and Palmer 2000; Keeling and Fast 2002; Keeling 2003). Based on phylogenetic analysis of small subunit ribosomal RNA as well as the lack of mitochondria, it has been suggested that microsporidia are one of the earliest eukaryote lineages (Leipe et al. 1993). More recent analysis indicate that microsporidia are related to fungi and not to early diverging eukaryotes (Hirt et al. 1999; Keeling, Luker, and Palmer 2000; Keeling 2003).
Similar to the fungi isoforms, the -actinin of E. cuniculi is short; it is only 537 amino acid residues long. The similarity to the prototypic -actinin is low (fig. 1). When submitted to the InterPro Scan and SMART, an actin-binding motif and two calponin homology domains in the N-terminal domain were identified. Further analysis of the primary structure indicated the highest similarity with sequences from fungi and amoebae (fig. 1).
Trichomonas vaginalis -Actinin
Trichomonas vaginalis is a sexually transmitted human parasite. When it adheres to the host cell, it undergoes a transformation from a flagellate to an amoeboid form. Previous analysis of the -actinin of T. vaginalis has suggested the presence of five repeats in the rod domain in addition to the common motifs in both termini (Addis et al. 1998; Bricheux et al. 1998). Interesting to note is that only the first of these repeats shows some similarity with other -actinins, and then only with the first repeat of other -actinins. It was suggested that the other four repeats have evolved because of intragenic duplication that has not occurred in other sequences (Bricheux et al. 1998).
Phylogenetic Analysis
The aligned sequences were analyzed using the PHYLIP, MEGA2, and MrBayes packages. The program SEQBOOT was used to create 1,000 bootstrapped data set and PROTPARS was used to calculate the phylogeny estimate. The maximum-parsimony tree obtained is shown in figure 2. Phylogenetic analysis based on neighbor-joining and Bayesian inference resulted in trees with very similar topologies as well as significance.
FIG. 2. Phylogenetic tree of -actinins. The most-parsimonious tree shows the relation between -actinins of different organism, as well as of different isoforms of the same organism. Bootstrap values in percentage were calculated from 1,000 data sets
There are three main branches in the tree; two vertebrate branches and one invertebrate and unicellular organisms branch. One of the vertebrate branches includes the calcium-insensitive isoforms 2 and 3, and the other branch includes the calcium-sensitive isoforms 1 and 4, thus giving rise to a phylogenetic tree with an ((A, B) (C, D)) topology. That -actinins of each isoforms are grouped together implies a higher degree of similarity between a particular isoform of all organisms than between the isoforms of a single organism. Most branches in the maximum-parsimony consensus tree are supported by bootstrap values of 90% or higher.
The phylogeny was also determined on a subset of the -actinin sequences. Analyzing the amino acid sequences of the actin-binding domain of the vertebrate -actinins resulted in very similar tree topologies. However, when the nucleotide sequences were used for the analysis instead, an (A, (B, (C, D))) topology was obtained.
In the invertebrate branch Ciona -actinin appears to be closest related to -actinins of the arthropod branch (Drosophila and Anopheles) and the nematode C. elegans. The placing of unicellular -actinins at the end of this branch indicates that these have branched off early in evolution.
Analysis of the Spectrin Repeats
We also analyzed the phylogenies of the spectrin repeats. All methods tested placed the single spectrin repeat of E. histolytica -actinins closest to the first spectrin repeat (SR1), as shown in figure 3. The two repeats of S. pombe and N. crassa are most similar to SR1 and SR4, respectively. This suggests a possible evolutionary pathway (fig. 4). In the primordial ancestor to present-day -actinins, the rod domain contained only a single spectrin repeat that became the first repeat (SR1) in modern -actinins. A first intragenic duplication gave rise to the second repeat, which became the fourth repeat (SR4) in modern -actinins. A subsequent second intragenic duplication added two more repeats (SR2 and SR3).
FIG. 3. Phylogenetic analysis of spectrin repeats in chicken, yeast, and amoeba. In the most-parsimonious tree, the single repeat of Enthamoeba histolytica (entamoeba) appears closest to the first repeat of chicken (SR1) -actinin. Similarly, the first repeat of the Schizosaccharomyces pombe isoform (pombe SR1) appears most related to the first repeat in the chicken isoform (SR1), whereas the second repeat of yeast -actinin (pombe SR2) seems to be closest to the fourth chicken repeat (SR4). The bootstrap values in percentage were calculated from 1,000 data sets
FIG. 4. Schematic structure of ancestral and modern -actinins. All -actinins have preserved the two calponin homology domains and, in most cases, the calcium-binding domain, whereas the rod domain differs significantly between different isoforms. Basal organisms, such as Entamoeba histolytica and Schizosaccharomyce pombe with one and two spectrin repeats, respectively, in the rod domain indicate that the two repeats in S. pombe and the four repeats present in other organisms are the result of two consecutive intragenic duplications from the single spectrin repeat of a primordial -actinin, like the one present in E. histolytica
Discussion
The existence of a notochord in the larvae stage implies that the urochordate Ciona share a common ancestor with vertebrates before the vertebrate radiation (Dehal et al. 2002; Gee 2002; Pennisi 2002). Since Ciona appears to have only a single -actinin gene in contrast to higher vertebrates such as human, mouse, and chicken, it seems likely that the appearance of the different -actinin isoforms arose after the vertebrate radiation.
This theory is substantiated by the topology of the phylogenetic tree, which indicated one invertebrate and two major vertebrate branches. One of the vertebrate branches encompasses the calcium-insensitive -actinin 2 and -actinin 3, and the other branch encompasses the calcium-sensitive isoforms 1 and 4. This ((A, B) (C, D)) tree topology lends support to the suggestion that two rounds of genome duplication occurred early in vertebrate evolution (the 2R hypothesis) (Hughes 1999; Durand 2003). However, the observed phylogeny can also be explained by independent duplication of individual genes or gene segments.
A phylogenetic analysis using the nucleotide sequences corresponding to the actin-binding domain of vertebrate -actinins resulted in a different tree topology (Dixson, Forstner, and Garcia 2003). In this case, the tree displayed a (A, (B, (C, D))) topology, suggesting that the calcium-insensitive -actinin 2 evolved initially, followed first by the calcium-insensitive -actinin 3 and then by the calcium-sensitive -actinins 1 and 4. When we repeated the phylogenetic analysis on this subset of sequence data, using the amino acid sequence of the actin-binding domain, we obtained tree topologies very similar to those obtained for the full data set. However, when we repeated the analysis using the nucleotide sequences, we obtained an (A, (B, (C, D))) tree topology, similar to that reported by Dixson, Forstner, and Garcia (2003). Thus, depending on whether amino acid or nucleotide sequences are used in the phylogenetic analysis, different tree topologies may be obtained.
It has been suggested that phylogenetic analysis based on amino acid sequences are more reliable than analysis based on nucleotide sequences because of compositional bias in the DNA sequences (Foster and Hickey 1999). This, together with the high bootstrap values, gives substantial support to our phylogenetic analysis.
Also frog (X. laevis) and fish (D. rerio and F. rubripes) appear to have a single -actinin isoform. The frog and puffer fish -actinins are related to isoform 1, whereas the -actinin of zebrafish is related to -actinin 3. The appearance of only a single isoforms in each of these organisms does not exclude the possibility of other isoforms, as their genomes have not yet been completely sequenced.
All tested phylogeny models placed the isoform of E. histolytica as the earliest diverging -actinin, followed by the -actinin of E. cuniculi. Together with the isoform of T. vaginalis, these two -actinins are also those that are least similar to the consensus sequence. Despite the low similarity, these proteins display certain of the typical hallmarks of -actinins and can be regarded as -actinins or at least as -actinin–like proteins.
The fungi S. pombe and N. crassa also branched off early in evolution. The fungi -actinins are more similar to the present day -actinins, despite the short rod domain with its two spectrin repeats.
When aligning the sequence of E. cuniculi -actinin with -actinins of E. histolytica and S. pombe, it is apparent that the sequence is similar not only to the S. pombe isoform but also to the E. histolytica -actinin, again placing microsporidia between the fungus and protozoa branches.
In all bacterial and plant genomes available to date, we have not been able to identify any genes for a -actinin or a -actinin–like protein. This implies that the primordial -actinin appeared in a primitive unicellular organism belonging to the protozoan kingdom.
-Actinin is composed of three major domains: an N-terminal actin-binding domain, consisting of two calponin homology domains; a rod domain of spectrin repeats, and a C-terminal calcium-binding domain. These domains are highly conserved throughout the animal kingdom. However, in basal eukaryotes and, in particular, in unicellular organisms, this domain structure is less well preserved. The evolutionary best-preserved domain is the N-terminal domain with its two calponin homology domains, which is present in all -actinins and -actinin–like proteins. This is not surprising, as the major function of -actinin is believed to be binding actin filaments (Blanchard, Ohanian, and Critchley 1989). Although it was not possible to identify a calcium-binding motif in all sequences, it is apparent that this domain is also well conserved among the different -actinins. It should be noticed that not all isoforms of -actinin are sensitive to calcium ions because of structural changes in this region during evolution (Dixson, Forstner, and Garcia 2003).
The most variable part is the rod domain, which is composed of one, two, or four spectrin repeats. In -actinins able to form antiparallel dimers, the function of the rod domain is mainly to separate the actin-binding domains of each molecule (Ylanne et al. 2001). Therefore, the rod part could be built from any amino acid sequence that can form a coiled-coil or rodlike structure. The -actinin of the protozoa Trichomonas vaginalis seems to support this conclusion, as only one spectrin repeat can be identified in the rod domain in addition to a long coiled-coil region. Also, the rod domain of E. histolytica -actinin contains only a single repeat unit, but in this case, there is no additional coiled-coil region.
The rod domain of the isoforms of E. histolytica and E. cuniculi are much shorter, and only a single spectrin repeat could be identified. Although the rod domain of T. vaginalis -actinin is longer, also in this case, only a single repeat could be identified.
The rod domain of -actinins of all vertebrates, as well as of invertebrates, has four spectrin repeats. This suggests a likely scenario for the evolution of the rod domain. In the primordial -actinin, the rod domain was short, spanning a single spectrin repeat. An intragenic duplication added a second repeat, followed by yet another intragenic duplication. As the single repeat in E. histolytica appears most related to the first repeat (SR1), it is likely that this repeat also was the first to evolve. Likewise, as the two repeats of fungi are closest to repeat 1 (SR1) and 4 (SR4), this suggests that repeat 4 aroused through intragenic duplication. The second intragenic duplication added repeats 2 (SR2) and 3 (SR3). Thus, our data do not support the conclusion that the ancestor of -actinin should possess four spectrin repeats (Viel 1999).
It has been demonstrated that SR2 and SR3 are essential for dimerization of -actinin (Djinovic-Carugo et al. 1999). In the dimer, SR2 and SR3 pair with the SR3 and SR2 of the other antiparallel molecule. Therefore, it is possible that these primitive -actinins are unable to form dimers, which would exclude any function involving actin cross-linking. The function of these short and atypical -actinins is still unknown.
Supplementary Material
The final alignment of all retrieved -actinin sequences is available online at www.mbe.oupjournals.org.
Acknowledgements
This work was supported by grants from EU HPRN-CT-2000–00096, ke Wibergs Stiftelse, Magn. Bergvalls Stiftelse and O.E. och Edla Johansson Stiftelse.
The TIGR Ciona intestinalis Gene Index updated February 2, 2004, now contains clone TC35693, which is identical to the Ciona -actinin sequence determined in this work.
Literature Cited
Addis, M. F., P. Rappelli, G. Delogu, F. Carta, P. Cappuccinelli, and P. L. Fiori. 1998. Cloning and molecular characterization of a cDNA clone coding for Trichomonas vaginalis alpha-actinin and intracellular localization of the protein. Infect. Immun. 66:4924-4931.
Bailey, G. B., P. S. Shen, M. J. Beanan, and N. E. McCoomer. 1992. Actin associated proteins of Entamoeba histolytica. Arch. Med. Res. 23:129-132.
Baines, A. J. 2003. Comprehensive analysis of all triple helical repeats in beta-spectrins reveals patterns of selective evolutionary conservation. Cell. Mol. Biol. Lett. 8:195-214.
Barstead, R. J., L. Kleiman, and R. H. Waterston. 1991. Cloning, sequencing and mapping of an alpha-actinin gene from the nematode Caenorhabditis Elegans. Cell Motil. Cytoskelet. 20:69-78.
Beggs, A. H., T. J. Byers, J. H. M. Knoll, F. M. Boyce, G. A. P. Bruns, and L. M. Kunkel. 1992. Cloning and characterization of two human skeletal muscle alpha-actinin genes located on chromosomes 1 and 11. J. Biol. Chem. 267:9281-9288.
Blanchard, A., V. Ohanian, and D. Critchley. 1989. The structure and function of alpha-actinin. J. Muscle Res. Cell Motil. 10:280-289.
Bricheux, G., G. Coffe, N. Pradel, and G. Brugerolle. 1998. Evidence for an uncommon -actinin protein in Trichomonas vaginalis. Mol. Biochem. Parasitol. 95:241-249.
Dehal, P., Y. Satou, and R. K. Campbell, et al. (87 co-authors). 2002. The draft genome of Ciona intestinalis: insights into chordate and vertebrate origins. Science. 298:2157-2167.
Dixson, J. D., M. J. Forstner, and D. M. Garcia. 2003. The -actinin gene family: A revised classification. J. Mol. Evol. 56:1-10.
Djinovic-Carugo, K., M. Gautel, J. Ylanne, and P. Young. 2002. The spectrin repeat: a structural platform for cytoskeletal protein assemblies. FEBS Lett. 513:119-123.
Djinovic-Carugo, K., P. Young, M. Gautel, and M. Saraste. 1999. Structure of the -actinin rod: molecular basis for cross-linking of actin filaments. Cell. 98:537-546.
Dubreuil, R. R. 1991. Structure and evolution of the actin crosslinking proteins. Bioessays. 13:219-226.
Durand, D. 2003. Vertebrate evolution: doubling and shuffling with a full deck. Trends Genet. 19:2-5.
Felsenstein, J. 1989. PHYLIP (phylogeny inference package) Version 3.2. Caldistics. 5:164-166.
Foster, P. G., and D. A. Hickey. 1999. Compositional bias may affect both DNA-based and protein-based phylogenetic reconstructions. J. Mol. Evol. 48:284-290.
Fyrberg, E., M. Kelly, E. Ball, C. Fyrberg, and M. C. Reedy. 1990. Molecular genetics of Drosophila alpha-actinin: mutant alleles disrupt Z disc integrity and muscle insertions. J. Cell Biol. 110:1999-2011.
Gee, H. 2002. Genomics: return of a little squirt. Nature. 420:755-756.
Germot, A., H. Philippe, and H. Le Guyader. 1997. Evidence for loss of mitochondria in Microsporidia from a mitochondrial-type HSP70 in Nosema locustae. Mol. Biochem. Parasitol. 87:159-168.
Hirt, R. P., J. M. Logsdon, Jr., B. Healy, M. W. Dorey, W. F. Doolittle, and T. M. Embley. 1999. Microsporidia are related to Fungi: evidence from the largest subunit of RNA polymerase II and other proteins. Proc. Natl. Acad. Sci. USA 96:580-585.
Huelsenbeck, J. P., and F. Ronquist. 2001. MRBAYES: Bayesian inference of phylogenetic trees. Bioinformatics 17:754-755.
Hughes, A. L. 1999. Phylogenies of developmentally important proteins do not support the hypothesis of two rounds of genome duplication early in vertebrate history. J. Mol. Evol. 48:565-576.
Janssen, K. P., L. Eichinger, P. A. Janmey, A. A. Noegel, M. Schliwa, W. Witke, and M. Schleicher. 1996. Viscoelastic properties of F-actin solutions in the presence of normal and mutated actin-binding proteins. Arch. Biochem. Biophys. 325:183-189.
Keeling, P. J. 2003. Congruent evidence from alpha-tubulin and beta-tubulin gene phylogenies for a zygomycete origin of microsporidia. Fungal. Genet. Biol. 38:298-309.
Keeling, P. J., and N. M. Fast. 2002. Microsporidia: biology and evolution of highly reduced intracellular parasites. Annu. Rev. Microbiol. 56:93-116.
Keeling, P. J., M. A. Luker, and J. D. Palmer. 2000. Evidence from beta-tubulin phylogeny that microsporidia evolved from within the fungi. Mol. Biol. Evol. 17:23-31.
Kumar, S., K. Tamura, I. B. Jakobsen, and M. Nei. 2001. MEGA2: molecular evolutionary genetics analysis software. Bioinformatics. 17:1244-1245.
Leipe, D. D., J. H. Gunderson, T. A. Nerad, and M. L. Sogin. 1993. Small subunit ribosomal RNA+ of Hexamita inflata and the quest for the first branch in the eukaryotic tree. Mol. Biochem. Parasitol. 59:41-48.
Mills, M., N. Yang, R. Weinberger, D. L. Vander Woude, A. H. Beggs, S. Easteal, and K. North. 2001. Differential expression of the actin-binding proteins, alpha-actinin-2 and -3, in different species: implications for the evolution of functional redundancy. Hum. Mol. Genet. 10:1335-1346.
Nickel, R., T. Jacobs, B. Urban, H. Scholze, H. Bruhn, and M. Leippe. 2000. Two novel calcium-binding proteins from cytoplasmic granules of the protozoan parasite Entamoeba histolytica. FEBS Lett. 486:112-116.
Norwood, F. L. M., A. J. Sutherland-Smith, N. H. Keep, and J. Kendrick-Jones. 2000. The structure of the N-terminal actin-binding domain of human dystrophin and how mutations in this domain may cause Duchenne or Becker muscular dystrophy. Struct. Fold Des. 8:481-491.
Pascual, J., J. Castresana, and M. Saraste. 1997. Evolution of the spectrin repeat. BioEssays. 19:811-817.
Pennisi, E. 2002. Comparative genomics: tunicate genome shows a little backbone. Science. 298:2111-2112.
Peyretaillade, E., V. Broussolle, P. Peyret, G. Metenier, M. Gouy, and C.P. Vivares. 1998. Microsporidia, amitochondrial protists, possess a 70-kDa heat shock protein gene of mitochondrial evolutionary origin. Mol. Biol. Evol. 15:683-689.
Roger, A. J., and J. D. Silberman. 2002. Cell evolution: mitochondria in hiding. Nature. 418:827-829.
Satou, Y., L. Yamada, and Y. Mochizuki, et al. (14 co-authors). 2002. A cDNA resource from the basal chordate Ciona intestinalis. Genesis. 33:153-154.
Tang, J., D. W. Taylor, and K. A. Taylor. 2001. The three-dimensional structure of alpha-actinin obtained by cryoelectron microscopy suggests a model for Ca2+-dependent actin binding. J. Mol. Biol. 310:845-858.
Thomas, G. H., E. C. Newbern, C. C. Korte, M. A. Bales, S. V. Muse, A. G. Clark, and D. P. Kiehart. 1997. Intragenic duplication and divergence in the spectrin superfamily of proteins. Mol. Biol. Evol. 14:1285-1295.
Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. 1997. The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 25:4876-4882.
Viel, A. 1999. -Actinin and spectrin structures: an unfolding family story. FEBS Lett. 460:391-394.
Weiss, L. M. 2001. Microsporidia: emerging pathogenic protists. Acta Trop. 78:89-102.
Williams, B. A., R. P. Hirt, J. M. Lucocq, and T. M. Embley. 2002. A mitochondrial remnant in the microsporidian Trachipleistophora hominis. Nature. 418:865-869.
Witke, W., A. Hofmann, B. Koeppel, M. Schleicher, and A. A. Noegel. 1993. The Ca2+-binding domains in non-muscle type -actinin: biochemical and genetic analysis. J. Cell Biol. 121:599-606.
Wu, J. Q., J. Bahler, and J. R. Pringle. 2001. Roles of a fimbrin and an -actinin-like protein in fission yeast cell polarization and cytokinesis. Mol. Biol. Cell. 12:1061-1077.
Ylanne, J., K. Scheffzek, P. Young, and M. Saraste. 2001. Crystal structure of the alpha-actinin rod reveals an extensive torsional twist. Structure (Camb). 9:597-604.(Ana Virel and Lars Backma)