Molecular Characterization and Evolution of Α-Actinin: from Protozoa to Vertebrates

Home , Actinin, Dystroglycan, Fimbrin, Gelsolin (cellular), MreB, Spectrin repeat

Molecular characterization and evolution of α-actinin: from protozoa to vertebrates

Ana Virel

Akademisk avhandling

Som med vederbörligt tillstånd av rektor vid Umeå universitet för avläggande av filosofie doktorsexamen vid teknisk–naturvetenskaplig fakultet kommer att offentligen försvaras på Kemiska institutionen i sal KB3A9, KBC, lördagen den 9 december 2006, kl. 10.00.

Fakultetsopponent: Dr. Kristina Djinovic-Carugo, Department for Biomolecular Structural Chemistry, Faculty of Chemistry, University of Vienna.

Department of Chemistry, Biochemistry, Umeå University Umeå 2006 Organization Document name UMEÅ UNIVERSITY DOCTORAL DISSERTATION Department of Chemistry Biochemistry SE-901 87 Umeå, Sweden Date of issue November 2006 Author Ana Virel

Title Molecular characterization and evolution of α-actinin: from protozoa to vertebrates

Abstract α-actinin is a ubiquitous protein found in most eukaryotic organisms. The ability to form dimers allows α-actinin to cross-link actin in different structures. In muscle cells α-actinin is found at the Z-disk of sarcomeres. In non-muscle cells α-actinin is found in zonula adherens or focal adhesion sites where it can bind actin to the plasma membrane. α-actinin is the shortest member of the spectrin superfamily of proteins which also includes spectrin, dystrophin and utrophin. Several hypotheses suggest that α-actinin is the ancestor of this superfamily. The structure of α-actinin in higher organisms has been well characterized consisting of three main domains: an N-terminal actin-binding domain with two calponin homology domains, a central rod domain with four spectrin repeats and a C-terminal calcium-binding domain. Data mining of genomes from diverse organisms has made possible the discovery of new and atypical α-actinin isoforms that have not been characterized yet. Invertebrates contain a single α-actinin isoform, whereas most of the vertebrates contain four. These four isoforms can be broadly classified in two groups, muscle isoforms and non-muscle isoforms. Muscle isoforms bind actin in a calcium- independent manner whereas non-muscle isoforms bind actin in a calcium-dependent manner. Some of the protozoa and fungi isoforms are atypical in that they contain fewer spectrin repeats in the rod domain. We have purified and characterized two ancestral α-actinins from the parasite Entamoeba histolytica. Our results show that despite the shorter rod domain they conserve the most important functions of modern α-actinin such as actin-bundling formation and calcium-binding regulation. Therefore it is suggested that they are genuine α-actinins. The phylogenetic tree of α-actinin shows that the four different α-actinin isoforms appeared after the vertebrate-invertebrate split as a result of two rounds of genome duplication. The atypical α-actinin isoforms are placed as the most divergent isoforms suggesting that they are ancestral isoforms. We also propose that the most ancestral α-actinin contained a single repeat in its rod domain. After a first intragene duplication α-actinin with two spectrin repeats were created and a second intragene duplication gave rise to modern α-actinins with four spectrin repeats.

Key words: α-actinin, evolution, spectrin repeat, intragene duplication, phylogenetic tree, spectrin superfamily.

Language: English ISBN: 91-7264-204-1 Number of pages: 52 + 4 papers Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Ana Virel

Department of Chemistry, Biochemistry, Umeå University Umeå 2006

Department of Chemistry, Biochemistry Umeå University SE-901 87 Umeå, Sweden

•

Organization Document name UMEÅ UNIVERSITY DOCTORAL DISSERTATION Department of Chemistry Biochemistry SE-901 87 Umeå, Sweden Date of issue December 2006 Author Ana Virel Title Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Abstract -actinin is a ubiquitous protein found in most eukaryotic organisms. The ability to form dimers allows -actinin to cross-link actin in different structures. In muscle cells -actinin is found at the Z-disk of sarcomeres. In non-muscle cells -actinin is found in zonula adherens or focal adhesion sites where it can bind actin to the plasma membrane. -actinin is the shortest member of the spectrin superfamily of proteins which also includes spectrin, dystrophin and utrophin. Several hypotheses suggest that -actinin is the ancestor of this superfamily. The structure of -actinin in higher organisms has been well characterized consisting of three main domains: an N-terminal actin-binding domain with two calponin homology domains, a central rod domain with four spectrin repeats and a C-terminal calcium-binding domain. Data mining of genomes from diverse organisms has made possible the discovery of new and atypical -actinin isoforms that have not been characterized yet. Invertebrates contain a single -actinin isoform, whereas most of the vertebrates contain four. These four isoforms can be broadly classified in two groups, muscle isoforms and non-muscle isoforms. Muscle isoforms bind actin in a calcium- independent manner whereas non-muscle isoforms bind actin in a calcium-dependent manner. Some of the protozoa and fungi isoforms are atypical in that they contain fewer spectrin repeats in the rod domain. We have purified and characterized two ancestral -actinins from the parasite Entamoeba histolytica. Our results show that despite the shorter rod domain they conserve the most important functions of modern -actinin such as actin-bundling formation and calcium-binding regulation. Therefore it is suggested that they are genuine -actinins. The phylogenetic tree of -actinin shows that the four different -actinin isoforms appeared after the vertebrate-invertebrate split as a result of two rounds of genome duplication. The atypical -actinin isoforms are placed as the most divergent isoforms suggesting that they are ancestral isoforms. We also propose that the most ancestral -actinin contained a single repeat in its rod domain. After a first intragene duplication -actinin with two spectrin repeats were created and a second intragene duplication gave rise to modern -actinins with four spectrin repeats.

Key words: -actinin, evolution, spectrin repeat, intragene duplication, phylogenetic tree, spectrin superfamily.

Language: English ISBN: 91-7264-204-1 Number of pages: 52 + 4 papers

The only true wisdom is in knowing you know nothing. •

Socrates

Why and how to study protein evolution?

Many times people ask me why do I study protein evolution? What is the important thing with it? Protein evolution is the study of how proteins have changed and evolved as a result of selection or specialization. The tools to study protein evolution are among others sequence comparison, structure prediction and phylogenetic trees. The publishing of diverse genomes give us the opportunity to identify proteins with unknown structure or function. By knowing the evolutionary story of different proteins we can predict the structure and function of these unknown proteins simply by comparison. The knowledge about function and structure of one protein will accelerate the understanding of related proteins. We can track evolution and predict how topology and functionality changed over time. We will be able to predict how mutations can affect protein structure, or which sequences are essential for the functionality of a particular domain. The study of protein evolution allows to recognize relations among proteins. Understanding how proteins evolve can be also useful to design effective and long lasting vaccines or drugs.

Prologue

When I first came to Umeå my project was to study the evolution of the spectrin superfamily of proteins. We decided to start studying -actinin since it was believed to be the ancestor of this superfamily. During this time I realized that little was known about the evolution of -actinin. At that point, I found myself immersed in a new unexplored world with lots of helpful information and uncharacterized -actinin isoforms. Now, 5 years later, I have collected in this thesis all my research about -actinin. The first part is a general introduction that shows the relevance of -actinin and its related proteins. The second part will discuss more in detail the structure, function and evolution of -actinin. Although my research has been more focus on the evolution of -actinin, it has also contributed to other relevant aspects associated with this protein.

•

List of Contents

LIST OF PAPERS ...... 1

ABREVIATIONS ...... 2

PART I INTRODUCTION ...... 3

1. THE CYTOSKELETON ...... 3

2. ACTIN ...... 4 2.1 Actin Structure ...... 5 2.2 Actin polymerization...... 5 3. ACTIN-BINDING PROTEINS ...... 8 3.1 Proteins regulating actin polymerization...... 8 3.2 Cytoskeletal linkers and membrane anchors...... 9 3.3 Cross-linking proteins ...... 9 4. SPECTRIN SUPERFAMILY...... 10 4.1 Spectrin superfamily characteristic domains...... 11 4.2 Members of the spectrin superfamily...... 12 4.3 Evolution of the spectrin superfamily ...... 15 PART II -ACTININ ...... 17

1. STRUCTURE...... 17 1.1 -actinin dimer...... 17 1.2 Actin-binding domain ...... 18 1.3 Rod domain...... 19 1.4 Calcium-binding domain: EF-hands ...... 19 2. -ACTININ ISOFORMS...... 20 2.1 Invertebrate isoforms ...... 20 2.2 Vertebrate isoforms...... 20 2.3 Expression of the different isoforms ...... 22 3. BIOLOGICAL FUNCTIONS OF -ACTININ...... 22 3.1 Actin cross-linking...... 23 3.2 Muscle -actinin...... 23 3.3 Non-muscle -actinin...... 24 4. REGULATORS OF -ACTININ...... 26 4.1 Regulation by calcium ...... 26 4.2 Regulation by phosphoinositides ...... 27

4.3 Regulation by phosphorylation...... 27 5. MOLECULAR EVOLUTION OF -ACTININ...... 28 5.1 Aim of the Study...... 28 5.2 Evolution of the -actinin family. Paper I ...... 28 5.3 Isolation and characterization of atypical -actinins from Entamoeba histolytica. Paper II and III ...... 31 5.4 Evolution of -actinin rod domain. Paper IV...... 34 6. FINAL CONCLUSIONS ...... 37

7. FUTURE PERSPECTIVES ...... 39

APPENDIX: PHYLOGENETIC ANALYSIS ...... 40

ACKNOWLEDGMENTS...... 43

REFERENCES ...... 45

•

Molecular characterization and evolution of -actinin: from protozoa to vertebrates

List of Papers

This thesis includes the following papers referred by their roman numerals:

I Virel A, Backman L (2004) Molecular evolution and structure of -actinin. Mol. Biol. Evol. 21:1024-31

II Virel A, Backman L (2006) Characterization of Entamoeba histolytica -actinin. Mol. Biochem. Parasitol. 145:11-7

III Virel A, Addario B, Backman L (2006) Characterization of Entamoeba histolytica -actinin2 (Submitted manuscript)

IV Virel A, Backman L (2006) A comparative and phylogenetic analysis of the -actinin rod domain (Submitted manuscript)

 1  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Abreviations

ABPs actin-binding proteins ABS actin-binding site ADP adenosine diphosphate

ATP adenosine triphosphate CBD calcium-binding domain

CH calponin homology F-actin filamentous actin G-actin globular actin

NMDA N-methyl-D-aspartate SR spectrin repeat

• SH3 Src homology 3

 2  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Part I Introduction

1. The cytoskeleton The skeleton of a vertebrate gives support to the tissues and plays an important role in body movement. One could say that eukaryote cytoskeleton is an analogue to this structure which acts as both a skeleton and muscle for the cells. The cytoskeleton is a meshwork of filaments that fill the cytoplasm of eukaryote cells. It is essential for cell stability, cell movement, and also for intracellular transport of different structures and organelles. There are three types of filaments responsible for the diverse activities of the cytoskeleton: microfilaments, microtubules and intermediate filaments (figure1). Microfilaments Microtubules

Centrosome

Nuclear laminar

Intermediate filaments Cell nucleus

Figure 1. Schematic representation of the cytoskeletal components within a cell.

Microfilaments Microfilaments are approximately 8 nm wide. They are composed of globular, monomeric actin (G-actin). Actin monomers are able to polymerize to form long and rigid filaments (F-actin). These filaments can be structured in loose networks or in tight bundles depending of the location. Microfilaments regulate cell shape and cell movement (dos Remedios et al. 2003).

 3  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Microtubules Microtubules are hollow structures with a diameter of approximately 25 nm. They are composed of polymers of - and -tubulin arranged in longitudinal rows called protofilaments. Each microtubule is generally composed of 13 protofilaments. They play an important role in cell division, cilia and flagella movements and they are also related with transport and location of organelles. Actin filaments and microtubules usually act together to give polarity to the cell (Wade et al. 1998). Intermediate filaments Intermediate filaments are structures of approximately 10 nm diameter. They can be connected to other filaments by proteins like plectin. Unlike microfilaments and microtubules, intermediate filaments are nonpolarized structures composed of a large number of different proteins. They provide structural support to the cells and are also important in cell migration. There is a special type of intermediate filaments called lamins that can be found in the nuclear lamina (Coulombe and Wong 2004). •

2. Actin Actin is one of the most abundant proteins in eukaryote cells, especially in muscle where it can reach 20% of the cell protein content. Although it was thought that actin is only present in eukaryotes, an actin-like protein called MreB has been found in prokaryotes and suggested to be the ancestor to eukaryote actins. This protein is 15% identical to other actins and has similar actin-like cytoskeletal roles (van den Ent et al. 2001). The amino acid sequence of actin is extremely conserved among different organisms exhibiting at least 85% sequence identity between different isoforms. Some lower eukaryotes such yeast contain a single actin gene, but higher eukaryotes contain more than one isoform. The different isoforms of this broad family are classified in three classes: - , - and -actin. -actin is primarily found in muscle whereas -actin and -actin are principally present in non-muscle cells. Actin is mainly located in the cytoplasm, but it can be also found in the nucleus. In muscle cells actin is associated with myosin and

 4  Molecular characterization and evolution of -actinin: from protozoa to vertebrates controls muscle contraction. In non-muscle cells it plays a variety of functions, such as change in cell shape, organelle transport, receptor mediated responses, and cell movement (dos Remedios et al. 2003).

2.1 Actin Structure Monomeric actin has a molecular weight of 43.000 Da and a size of 67×40×37 Å. About 40% of the structure is -helix, but it also contains some -sheets (Figure 2) (dos Remedios et al. 2003). Actin can be divided into four subdomains. ATP is bound in the center of the molecule, where the four subdomains meet. The conformational change between the ADP- and ATP-actin leads to the reorientations of the different subdomains with respect to each other. There are five secondary cation binding sites on the actin surface suggested to play an important role in the actin polymerization (Estes et al. 1992; Otterbein et al. 2001) (figure 2).

Figure 2. Actin monomer ribbon structure of rabbit skeletal muscle (Otterbein et al. 2001) Pdb accession number: 1J6Z.

2.2 Actin polymerization Actin can polymerize into helical filaments under physiological conditions of neutral pH and high concentrations of K+ and Mg+2. The process of polymerization consists in three phases: first, the slow formation of a dimer that tends to dissociate rather than to assemble. Second, the formation of a trimer, consisting of a stable nucleation structure, and third, the elongation phase where the monomers are rapidly assembled (figure 3).

 5  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

+ - ADP-actin ATP-actin

Figure 3. Schematic overview of actin polymerization.

2.1.1 Actin treadmilling Actin filaments are polarized; they have a fast-growing (barbed end) and a slow-growing end (pointed end). Actin has ATPase activity although ATP is not required for the polymerization process. When G-actin has bound ATP it is incorporated to the barbed end of the filament. Subsequently ATP-actin in the filament is hydrolyzed to ADP-actin and shifted towards the pointed end. This directional growing of F-actin is called treadmilling. Actin filaments are in a continuously assembly/disassembly cycle that can be divided in three • parts: i) depolymerization and release of G-ADP-actin at the pointed-end, ii) exchange of ATP for ADP bound to actin, iii) addition of G-ATP-actin at the barbed end. In the cell the barbed end is oriented towards the membrane while the pointed end is in opposite direction. Treadmilling is used like a motor for cell motility as well as for pathogen locomotion (Disanza et al. 2005).

2.2.2 How to study actin polymerization Actin represents about 20% of the total protein content in muscle, therefore striated muscle is the most widely used source to purify actin for biochemical and biological studies. The most common procedure to extract actin was already used half a century ago. Minced muscle is dehydrated with acetone to denature unwanted proteins. After a series of polymerization/depolymerization steps by adding high and low ionic strength buffers, and subsequently centrifugations, actin is extracted with a >90% purity (Spudich and Watt 1971).

 6  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Actin polymerization can be studied by different methods:

DNase I assay.y Monomeric actin strongly binds deoxyribonuclease I and inhibits its endonuclease activity. This permits the measure of the levels of unpolymerised and filamentous actin (Lazarides and Lindberg 1974). Viscosity assay. Long actin filaments in solution increase viscosity and change the mechanical properties of the actin solution. Changes in F-actin viscosity under different conditions or in presence of certain proteins can be measured by the use of viscometers (Zaner and Stossel 1982). Pyrene-actin assay. Actin has a cysteine at the N-terminus that can be labelled with the aromatic hydrocarbon pyrene. Pyrene can be excited at 365 nm and the emission can be followed at 407 nm. When pyrenyl-actin is polymerized the florescence signal increases considerably (Cooper et al. 1983). The time course of actin polymerization and its different phases are easy to follow with this method. A typical curve of actin polymerization is showed in figure 5. The kinetics of the polymerization and therefore the characteristics of the curve can change due to different conditions or in presence of other proteins (Cooper et al. 1983).

Fluorescence

Time

Lag phase Growth phase Steady state

Figure 5. Time course of actin polymerization. The lag phase is due to the rate-limiting nucleation. The growth phase is when monomers are added to the ends of the polymer. The steady state is when addition of monomers is in equilibrium with the dissociation of monomers.

 7  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Electron microscopy.y Negatively stained actin filaments can be examined by electron microscopy to observe how the filaments are ordered in presence or absence of different proteins. Centrifugation. At high speed centrifugation (typically 100,000 rpm for 1h), actin filaments sediment leaving unpolymerized monomers in solution. At low speed centrifugation (13,000 rpm for 15 min) actin filaments alone do not sediment. However, in the presence of a cross-linking protein, actin filaments will sediment together with the protein. This method is widely used to study if a protein binds or cross-links actin filaments.

3. Actin-binding proteins Actin does not only interact with itself to form polymers, it also interacts with many other proteins that help and regulate all these important processes. Without these actin-binding proteins (ABPs), the • rapid actin turnover necessary for example for locomotion would be impossible. Although a large number of ABPs have been identified and characterized, it is very probable that many others still await discovery.

3.1 Proteins regulating actin polymerization The formation of the stable nucleus at the starting of the polymerization process is kinetically unfavourable. In the cell there are proteins that speed up this process. For example the Arp2/3 complex nucleates actin on a existing actin filament generating actin branches (Winder and Ayscough 2005). Filament growth is regulated by capping proteins like gelsolin and tensin. They can bind to the barbed ends of the filaments thereby blocking the addition of new monomers. Gelsolin can also sever actin filaments, increasing the rate for the filament disassembly (dos Remedios et al. 2003; McGough 1998). Other examples of severing proteins are actin depolymerization factor (ADP), and cofilin. Another group of ABPs like tropomyosins and nebulins can stabilize actin filaments by binding along the actin filament to protect against depolymerization (Winder and Ayscough 2005).

 8  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

The rapid response to intracellular and extracellular signals regulating the actin filament growth requires a continuous availability of actin monomers. Monomer binding proteins like profilin and thymosin bind G-actin as it is released from filament ends and facilitate the exchange of ADP for ATP (Perelroizen et al. 1996). There are also proteins involved in the signalling and regulation of the actin cytoskeleton, some examples are: GTPases such as Rho and Rac (Burridge and Wennerberg 2004).

3.2 Cytoskeletal linkers and membrane anchors Actin is mainly concentrated at the cell cortex close to the plasma membrane. There are many proteins such as dystrophin, talin and vinculin that connect actin directly to transmebrane proteins like integrins or cell membrane receptors such as dystroglycan. Other proteins as plectin can attach actin to microtubules and intermediate filaments (Winder and Ayscough 2005).

3.3 Cross-linking proteins Actin filaments can be arranged into higher-ordered, cross-linked structures. Cross-linking proteins require two actin-binding sites in order to bind two different actin filaments. They determine the different actin arrays as well as the distances between two actin filaments. There are many different cross-linking proteins; the most ubiquitous are -actinin, filamin, spectrin and fimbrin (Dubreuil 1991), these proteins can arrange cortical actin filaments in different structures (Figure 6): Parallel bundles are closely and parallel oriented, they are formed by proteins such as fimbrin and can be found in filopodia (Small et al. 2002). Contractile bundles are placed in antiparallel manner and are more loosely spaced, they are found in stress fibers together with proteins such as -actinin (Katoh et al. 1998). Gel-like networks are found for example in the membrane of erythrocytes where spectrin cross-links short actin fragments, generating hexagonal structures (Liu et al. 1987).

 9  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Filopodium

Cell cortex S tress fibers Fimbrin

ZZZZZZ ZZZZZZ ZZZZ ZZZZ ZZZZ

ZZZZZZ ZZZZZZ ZZ ZZZZZZ ZZZZ -actinin Filam in Calcium binding domain Actin binding domain -helical motif ZZZZZZZ -sheet repeats

• Figure 6. Cortical arrays of actin filaments in a crawling cell. Different cross-linking proteins arrange actin filaments in a different manner. The black arrows points towards the plus end of the actin filaments.

4. Spectrin superfamily The spectrin superfamily of proteins is constituted principally by spectrin, dystrophin, utrophin and -actinin (Dubreuil 1991). They share some common domains and act as actin-bundling or membrane-anchoring proteins in the cell. Since my main interest of study is -actinin I will explain with more detail different aspects of this superfamily that will help to better understand the structure and evolution of -actinin.

 10  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

4.1 Spectrin superfamily characteristic domains

4.1.1 Actin-binding domain (ABD) The actin-binding domain of the members of the spectrin superfamily consists of two tandem helical domains of about 125 amino acids. Each domain is described as a calponin homology (CH) domain because its close similarity to the N-terminus of calponin (Castresana and Saraste 1995). The two CH domains appear to have similar structure but differ in functionality (Gimona et al. 2002).

4.1.2 Calcium-binding domain (CBD) The calcium-binding domain in the members of the spectrin superfamily shows homology with calmodulin (Liu et al. 2004; Trave et al. 1995). This domain contains EF-hands motifs that consist of a helix-loop-helix structure that binds one calcium molecule. Some spectrin or -actinin isoforms are able to bind calcium whereas it has been suggested that dystrophin and utrophin do not. The binding of calcium to the EF-hands induces a conformational change able to regulate the actin-binding activity of the adjacent domain (Lundberg et al. 1992; Trave et al. 1995)

4.1.3 Rod domain: spectrin repeats The rod domain constitutes the central part of the isoforms and links the actin-binding domain with the C-terminal calcium-binding domain. It is composed of several repeats described as spectrin repeats because they were initially discovered in spectrin (Speicher and Marchesi 1984). The number and composition of the spectrin repeats varies among the different members, but their general structure is very similar (Djinovic-Carugo et al. 2002). Spectrin repeats are independent folding units composed of a left-handed antiparallel triple helical bundle where the three -helices wrap around each other. There is a heptad pattern of hydrophobic and polar residues that are essential for the fold of the repeat (Djinovic-Carugo et al. 2002; Parry et al. 1992). The length of the repeats varies slightly, in -actinin it is around 114-125 (Gilmore et al. 1994) and in spectrin and dystrophin the repeats are 106 and 109, respectively (Parry et al. 1992). It is possible to recognize some conserved residues in all the repeats (Pascual et al. 1997). There is a conserved tryptophan in helix A, a proline residue in

 11  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

helix B, an aromatic residue in helix C and a leucine close to the very C-terminal of each repeat. The number of spectrin repeats in the rod domain can vary from 2 to 4 in -actinin (Paper I; IV) to 17 and 24 repeats in spectrin and dystrophin respectively (Broderick and Winder 2005). Spectrin repeats can be found in cell structures under mechanical stress, such as cell cortex, Z-disk, and stress fibers. It also serves as a docking surface for other proteins. The rod domain in spectrin and dystrophin has been suggested to contribute to actin-binding (Amann et al. 1998; Li and Bennett 1996). -actinin and spectrin forms dimers (Djinovic-Carugo et al. 1999) whereas there is no evidence yet of dimer formation in dystrophin.

4.2 Members of the spectrin superfamily Spectrin was the first member where the characteristic triple-helical repeats where identified. Later, these repeats were also recognized in other related proteins such as -actinin and dystrophin (Davison et al. 1989) constituting the spectrin superfamily of proteins. Although these • proteins have different functions, all members are able to bind actin filaments. They also share similar structure and conserved domains (figure 7). As mentioned above, these proteins are characterized by having an N-terminal actin-binding domain (ABD), a C-terminal calcium-binding domain (CBD) and a central rod domain.

4.2.1 -actinin -actinin is composed of an N-terminal ABD with two calponin homology domains, a central rod domain with four spectrin repeats and a calcium-binding domain at the C-terminus. Recently it has been shown that some ancestral -actinins contain one or two spectrin repeats in the central rod domain (Paper I; IV). The function of -actinin is to cross-link actin filaments in different cell structures (Blanchard et al. 1989). There is evidence of -actinin in metazoan, protozoa and yeast but not in plants (Paper I).

4.2.2 Dystrophin and Utrophin Dystrophin appears to be present only in metazoan (Roberts and Bobrow 1998) where it plays an important role in binding the actin cytoskeleton to the extracelullar matrix (Ervasti and Campbell 1993). Utrophin has been considered the autosomal homolog of dystrophin

 12  Molecular characterization and evolution of -actinin: from protozoa to vertebrates due to the high structural and functional similarity (Winder et al. 1995b). Both proteins contain an ABD at the N-terminus and a long rod domain with 24 and 22 spectrin repeats in dystrophin and utrophin, respectively. The C-terminal contains a cysteine-rich domain with EF-hands motifs (Koenig et al. 1988; Winder et al. 1995a), a WW motif that binds proline-rich sequences, and a zinc finger motif involved in protein-protein interactions. The last residues of the C-terminus form a coiled-coil structure that can interacts with other dystrophin-related proteins (Broderick and Winder 2005). Mutations in the dystrophin gene located in the chromosome X can lead to Duchenne muscular dystrophy (Koenig et al. 1987).

4.2.3 Spectrin Spectrin was first discovered as a major component of the erythrocyte membrane skeleton. In the erythrocyte spectrin binds short actin filaments creating a structure that is essential for cell shape and membrane stability. In humans the two different isoforms, - and -spectrin, are encoded by at least 2 or 5 genes, respectively (Bennett and Baines 2001). -spectrin contains one partial repeat at the N-terminus followed by 20 complete spectrin repeats. Between spectrin repeat 9 and 10 there is a SH3 (src homology 3) domain. The C-terminal contains EF-hands motifs and is related with calmodulin (Trave et al. 1995). -spectrin has two tandem CH domains at the N-terminus that bind actin. The central rod domain is composed of 17 spectrin repeats. Repeat 17 is a partial repeat that interacts with the N-terminus of -spectrin to form a complete repeat. In some isoforms there is a pleckstrin homology domain (PH) at the very C-terminus suggested to be an important ligand-binding site. The high molecular weight H-spectrin is found in Drosophila melanogaster and Caenorhabditis elegans there is also an ortholog isoform in humans. H-spectrin contains 30 spectrin repeats instead of 17 and a SH3 motif inserted in the fifth triple helical repeat (Bennett and Baines 2001).

In the spectrin heterodimer, - and -subunits interacts laterally and in antiparallel manner (figure 7). Two heterodimers associate head-to-head to form a tetramer: the N-terminus of -spectrin associates with the C-terminus of -spectrin of the other heterodimer (Bennett and Baines 2001). Although there is evidence of

 13  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

immunoreaction against spectrin analogues in protozoa (Bennett and Condeelis 1988; Pollard 1984) we have not found any spectrin-like gene when searching in the completed genomes of available protozoas. Thus, we believed that spectrin is a characteristic of metazoan (Bennett and Baines 2001).

-actinin

CH1CH2 EFh

Dystrophin CH1CH2 W EFh ZZ

H-spectrin CH1CH2

-spectrin • CH1CH2 EFh -spectrin

Spectrin repeat CH1CH2 Actin binding domain W WW domain SH3 domain EFh EF-hand motif Coiled-coils PH domain ZZ Zinc finger motif

Figure 7. Schematic representation of the different members of the spectrin superfamily (Adapted from Broderick and Winder 2005).

 14  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

4.3 Evolution of the spectrin superfamily Phylogenetic and comparative analyses have revealed that the different members of the spectrin superfamily arose after several intragene duplications and rearrangements from an ancestral -actinin (Byers et al. 1989; Thomas et al. 1997; Wasenius et al. 1987). The similarity of the spectrin repeats present in -actinin (ac1-ac4), -spectrin (a0-a20) and -spectrin (b1-b17) has made possible to follow the evolution of this superfamily. The phylogeny shows that the two first repeats in -actinin (ac1 and ac2) share a common ancestor with the two first repeats in -spectrin (b1 and b2). The two last repeats in -actinin (a3-a4) share a common ancestor with the two last repeats in -spectrin (a19 and a20) (Byers et al. 1989; Dubreuil et al. 1989). It is known that the four repeats in -actinin are involved in dimerization (Djinovic-Carugo et al. 1999). Additionally, it has been shown that in the spectrin tetramer the repeats involved in dimerization are those related to -actinin. The other repeats found in - and -spectrin are more related to each other than to any repeat in -actinin, suggesting the existence of a common ancestor for - and -spectrin. This hypothesis is also supported by the finding of a heavy spectrin isoform containing typical domains of both - and -spectrin (Thomas et al. 1997). The repeats in dystrophin and utrophin are the most divergent and difficult to compare probably because they diverged earlier in evolution (Pascual et al. 1997). The most likely scenario for the evolution of the spectrin superfamily of proteins (figure 8) began with the introduction of seven repeats between repeat 2 and 3 of -actinin producing an elongated “-actinin”. At this point the dystrophin/utrophin lineage presumable diverged from the - -spectrin lineage. The next step was the duplication of the seven-repeat block in the elongated “-actinin” and the insertion of the ancestor of the tetramerization repeat (b17/a0). The tetramerization repeat was then split into b17 and a0 giving rise to the ancestral - and -spectrin subunits. The ancestral -spectrin then evolved by an elongation process. The ancestral -spectrin evolved, by a second duplication and insertion of single repeats (Pascual et al. 1997). The evolution of this superfamily can be divided in two phases. A first active phase characterized by intragene duplication and concerted evolution. A selection pressure favouring or disfavouring different

 15  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

protein lengths ended with the prevention of the array exchanges producing dystrophin and spectrin lineages. In a second stable phase the number of repeats became constant and their sequences evolved independently (Thomas et al. 1997).

ABD R1 R2 R3 R4 EFh -actinin Elongation

unstable lineage Dystrophin/ Ancestral Utrophin block Duplication /-spectrin ancestor Chopping

ancestral ancestral -spectrin Duplication • Elongation - spectrins

Elongation 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Modern -spectrin 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20

Modern -spectrin

Figure 8. Evolution of the spectrin superfamily. The red spheres denote -actinin repeats involved in dimerization. The black spheres denote spectrin repeats that were involved in duplication or elongation events (adapted from Pascual et al. 1997; Thomas et al. 1997).

 16  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Part II -actinin

1. Structure 1.1 -actinin dimer -actinin is composed of an N-terminal ABD, a central rod domain and a C-terminal calcium-binding domain. -actinin exists in a dimeric form where the two monomers interact in an antiparallel manner (Blanchard et al. 1989). The major dimer interface is in the rod domain creating a rigid and long structure (Ylanne et al. 2001) (figure 9). As a result, the two ABD are placed at each end of the dimer and in close vicinity with the calcium-binding domain of the other monomer. The antiparallel position of the two ABD allows -actinin to cross-link two adjacent actin filaments.

A BD S R 2 S R 1 CBD S R 3 S R 4

S R 4 S R 2 S R 3 CBD S R 1 A BD

Figure 9. Cryoelectron microscopy structure of chicken gizzard smooth muscle. (Pdb accession number: 1sjj)

ABD - actin-binding domain SR - spectrin repeats CBD - Calcium-binding domain

 17  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

1.2 Actin-binding domain -actinin belongs to a group of F-actin cross-linking proteins in which the actin-binding domain consists of two tandem CH domains (Gimona et al. 2002). Actin-binding studies of the two independent CH domains have shown that although the two domains are structurally similar they have different roles in actin-binding. CH1 domain alone binds actin with lower affinity than the entire ABD and the CH2 domain alone does not bind actin at all. This suggests that both domains are necessary for the maximum binding capacity of the entire domain (Banuelos et al. 1998; Way et al. 1992). Recently, the crystal structures of the ABD from human -actinin-3 and -actinin-1 were solved (Borrego-Diaz et al. 2006; Franzot et al. 2005). In both structures the CH1 and CH2 are in extensive contact and create a compact structure described as “close conformation”. Each CH domain is composed of two minor helices (B and F) and four principal -helices (A, C, E and G) that form the core of the domain. A flexible linker connects the two CH domains (Franzot et al. 2005). In the complete ABD there are three different binding sites for F-actin • called ABS1-3. ABS1 is placed in the first N-terminal helix of CH1, ABS2 is in the C-terminal helix of CH1, and the ABS3 comprises the interdomain linker and the N-terminal helix of CH2 domain. In the close conformation structure ABS2 and ABS3 are contiguously placed on the surface while ABS1 is in the interface of the two CH domains. It is therefore suggested that when the ABD is bound to actin it would experience a conformational change where the interaction between CH1-CH2 is disrupted and the ABS1 is more exposed to actin (Borrego-Diaz et al. 2006; Franzot et al. 2005). Cryoelectron microscopy and image analysis have suggested that -actinin binds F-actin subdomains 1 and 2 and subdomain 1 of the adjacent actin monomer (McGough et al. 1994). The amino acid residues involved in the binding are mainly hydrophobic, suggesting that the binding to actin is mediated by hydrophobic interactions (Kuhlman et al. 1992).

 18  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

1.3 Rod domain The rod domain of -actinin contains the characteristic spectrin repeats and is considered the shortest among the members of the superfamily. In higher eukaryotes -actinin comprises for repeats, however it has been found that some protozoa and fungi present -actinins with one or two putative spectrin repeat-like structures (paper I ; IV). The rod domain of -actinin from human and the two first spectrin repeats of Dictyostelium discoideum -actinin have been crystallized and studied into detail (Djinovic-Carugo et al. 1999; Kliche et al. 2001; Ylanne et al. 2001). The repeats present in -actinin with four repeats (SR1-SR4) have the conserved residues and structure of typical spectrin repeats. The major differences between the repeats are in the loops connecting the helices (Ylanne et al. 2001). The monomeric rod presents a slight change in charge, from a positive N-terminus to a more negative C-terminus (Djinovic-Carugo et al. 1999; Ylanne et al. 2001). The repeats have different interface properties that allow them to interact specifically with repeats from the other monomer giving rise to a rigid and unbendable rod domain. Specifically, the SR1 and SR2 interact with SR4 and SR3 respectively (Djinovic-Carugo et al. 1999; Flood et al. 1997). The linker between the repeats is helical and has no apparent break (Djinovic-Carugo et al. 1999). The most important characteristics of the entire rod domain are the overall acidic surface and the left-handed twist from one end to the other resulting in a structure where the ends are twisted 90°. It is believed that the acidic concave surface could represent a high affinity binding site for other proteins (Ylanne et al. 2001).

1.4 Calcium-binding domain: EF-hands The C-terminal calcium-binding domain of -actinin contains four EF-hands motifs that are related to calmodulin (Tang et al. 2001). The two first EF-hands are clearly recognized by sequence analysis while the other two are cryptic. The structure of the EF-hands 3 and 4 from human -actinin-2 has been solved by NMR in complex with the seventh Z-repeat of titin (Atkinson et al. 2001). Sequence analyses of the four EF-hands have indicated that EF-hands 3 and 4 present an amino acid residue deletion and insertion, respectively. The first two

 19  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

EF-hands are therefore the only with potential affinity for calcium (Tang et al. 2001). -actinin isoforms can be divided in two classes depending on their calcium-binding properties. Non-muscle isoforms contain calcium-binding EF-hands that regulate the actin-binding activity (Noegel et al. 1987; Witke et al. 1993). Mutagenesis studies of the two first EF-hands of D. discoideum -actinin have shown that the EF-hand 2 has a high-affinity calcium site, while EF-hand 1 has a lower affinity and acts as a regulatory calcium site for the actin-binding inhibition (Witke et al. 1993). Muscle isoforms contain non-functional EF-hands, they have lost the essential residues to coordinate calcium and therefore their capacity to bind actin is not regulated by calcium (Blanchard et al. 1989).

2. -actinin isoforms • 2.1 Invertebrate isoforms Invertebrates such as Drosophila melanogaster and Caenorhabditis elegans contain a single -actinin gene (Barstead et al. 1991; Roulier et al. 1992). In Drosophila, it has been observed that the single -actinin gene can be alternatively spliced giving three different isoforms that are expressed differently (Figure 10). The splicing occurs at the neck region between the ABD and the first spectrin repeat (Roulier et al. 1992). There is also evidence of atypical -actinin isoforms in some protozoa and yeast (papers I; II; III; IV; Addis et al. 1998; Bricheux et al. 1998; Wu et al. 2001). Most of these isoforms are shorter and contain one or two putative spectrin repeats in the rod domain.

2.2 Vertebrate isoforms Almost all vertebrates appear to have four different -actinin genes (ACTN1-4) (paper I; MacArthur and North 2004). It is noteworthy that to date no ACTN3 gene has been found in chicken. The different -actinin isoforms can be broadly classified in two groups: muscle -actinins which bind actin in a calcium-independent manner and non-muscle isoforms which bind actin in a calcium-dependent manner.

 20  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

-actinin-1 and -actinin-4 are generally the most abundant isoforms in non-muscle cells, whereas -actinin-2 and -actinin-3 are generally found in muscle tissue. Alternative splicing can give rise to additional diversity in some vertebrates (figure 10). The diversity is generated by the use of two different exons, one containing an active EF-hand region present in non-muscle -actinins, and the other present in muscle -actinin containing a deletion between the two non-functional EF-hands. The alternate use of these two exons has been seen in chicken -actinin-2 (Parr et al. 1992) and in chicken and rat -actinin-1 (Kremerskothen et al. 2002; Waites et al. 1992). A particular brain isoform of -actinin-1 containing both exons has been found in rat (Kremerskothen et al. 2002). Recently a splice variant of -actinin-4 with higher affinity for actin has been related with lung and testis cancer. The alternative spliced exon contains part of the linker between the ABD and the first spectrin repeat (Honda et al. 2004).

A BD SR1 SR2 SR3 SR1 CBD invertebrates N on-m uscle A dult m uscle Larval m uscle

A BD SR1 SR2 SR3 SR1 CBD vertebrates actn4 actn1 actn2 Cancer cells N on-m uscle EFEF S arcom eric EF EF S m ooth m uscle EF EF N on-m uscle EF EF N on-m uscle

EF EF Brain Spacer No spacer

Figure 10. Alternative splicing of vertebrate and invertebrate -actinins. Within the Boxes are represented the alternative exons used to give rise to the different isoforms. (Adapted from MacArthur and North 2004). ABD - actin-binding domain; SR - spectrin repeats; CBD – calcium-binding domain

 21  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

2.3 Expression of the different isoforms

2.3.1 -actinin-1 and -actinin-4 Both isoforms are present in stress fibers and lamellipodia. -actinin-1 is more abundant than -actinin-4 and is found in cell adhesions structures. -actinin-4 is involved in cell movement and is found in the nucleus (Honda et al. 1998) and significantly expressed in kidney (Kos et al. 2003) .

2.3.2 -actinin-2 and -actinin-3 They are almost exclusively expressed in muscle cells. Human and mouse -actinin-2 are expressed in skeletal and cardiac muscle, whereas -actinin-3 is limited to skeletal muscle (Beggs et al. 1992; Mills et al. 2001). Surprisingly there is no -actinin-3 in chicken. -actinin-2 and -actinin-3 appears to have overlapping function in human (Mills et al. 2001).

• 3. Biological functions of -actinin Much is known about -actinin function and interacting partners. The list of newly identified proteins and molecules interacting with -actinin has increased during the last years, which confirms that -actinin is not merely an actin-binding protein but a scaffold for other protein-protein interactions and regulations. -actinin was first isolated from striated muscle cells (Ebashi and Ebashi 1965; Maruyama and Ebashi 1965). Later, -actinin was also localized by inmunofluorescence along actin stress fibers, cell substrate and cell-cell attachments in non-muscle cells (Lazarides and Burridge 1975). I will briefly mention some of the most outstanding functions and interactions where -actinin is involved.

 22  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

3.1 Actin cross-linking The highly conserved ABD in -actinin suggests that binding to actin is one of the most important features of -actinin. When -actinin binds actin it can form parallel or antiparallel bundles (Meyer and Aebi 1990; Taylor et al. 2000). The manner of how -actinin is able to form parallel or antiparallel bundles is not yet well understood. The most plausible mechanism is that the neck between the ABD domain and the rod domain allows flexibility and reorientation of the actin-binding sites (Winkler et al. 1997). It has been shown that different -actinin isoforms differ in their affinity for actin and that the bundle formation critically depend of the concentration ratio of -actinin and actin (Meyer and Aebi 1990).

3.2 Muscle -actinin -actinin plays an important role in all muscle tissues (Endo and Masaki 1982; Fay et al. 1983; Lemanski et al. 1980; Suzuki et al. 1976). The most common feature of the different muscle isoforms is the binding to actin in a calcium-independent manner (Blanchard et al. 1989). In striated muscle -actinin is found at the Z-disk, attaching actin filaments of adjacent sarcomeres (figure 11). The periodicity of -actinin in these structures determines the thickness of the Z-disk (Luther 1991). Titin has been shown to mediate the assembly of -actinin to the Z-disk. Titin extends from the Z-line to the M-line and has two interaction sites for -actinin (Young et al. 1998). Myotilin and myopaladin belong to a novel family of sarcomeric protein, both bind to -actinin in the Z disk and are essential for the normal muscle structure (Bang et al. 2001; Salmikangas et al. 1999).

 23  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

A band Z line M line I band

myosin ααα-actinin F-actin

Titin

• Figure 11. Scheme of a sarcomere under electron microscope. The M - line corresponds to the location where different myosin filaments overlap. The Z - lines are at either ends of the sarcomeres. Proteins like -actinin or titin are the responsible for the attachment of different actin filaments.

3.3 Non-muscle -actinin

3.3.1 Adhesion sites -actinin can be found in adhesion sites such as cell-cell contact, neuronal synapses and focal adhesion sites (Otey and Carpen 2004) (figure 12). In those structures -actinin is bound to the cytoplasmic face of different adhesion and transmembrane proteins, like for instance catenin (Knudsen et al. 1995), integrin (Otey et al. 1990) or vinculin (Belkin and Koteliansky 1987). These interactions are generally mediated by the acidic rod domain surface that links the actin cytoskeleton to the plasma membrane providing structural stability. -actinin can also gather different signalling molecules suggesting its participation in different signalling pathways between transmembrane proteins and actin filaments (Otey and Carpen 2004).

 24  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

FOCAL ADHESION SITE

actin ααα-actinin α α V α α V

V V talin vinculin V α α talin α α V

plasma βββ ααα integrin βββ membrane ααα βββ ααα extracellular matrix

Figure 12. Schematic representation of -actinin and its interacting partners in a focal adhesion site.

3.3.2 Cell Migration Several studies have shown that -actinin is involved in cell migration. The precise mechanism is not well understood, but it is believed that -actinin is involved in the step of de-adhesion (Otey and Carpen 2004). In adhesion plaques modulation in the expression of -actinin affect the stability of the structure leading to increase or decrease of cell attachment (Gluck and Ben-Ze'ev 1994).

3.3.3 Stress fibers Stress fibers are bundles of actin filaments characterized by their ability to support tension (Katoh et al. 1998). Dense bodies are spread along stress fibers and are suggested to be the non-muscle analogous of Z-disk structures. -actinin is a component of these dense regions (Lazarides and Burridge 1975). There are several molecules that bind -actinin in dense bodies regions such as PDZ- and LIM-domain proteins (Otey and Carpen 2004).

3.3.4 Synapses It has been shown that -actinin-2 binds to the neurotransmitter receptor NMDA, the precise function is not yet known, but it seems

 25  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

that -actinin serves as an anchor for NMDA to the actin cytoskeleton (Wyszynski et al. 1997). -actinin-4 binds to the cytoplasmic PDZ-domain of the transmembrane protein densin in synapses (Walikonis et al. 2001). It is suggested that these complex might contribute to the stability of the synaptic cell junction by calcium regulation (Otey and Carpen 2004).

3.3.5 Cleavage furrow Fluorescence analysis has shown that -actinin is concentrated in the contractile ring creating the cleavage furrow during cytokinesis. The results suggested that -actinin and other cytoplasm proteins might generate the force necessary for the partition of two daughter cells (Fujiwara et al. 1978).

3.3.6 Cancer and metastasis Metastasis is the movement or spread of carcinogenic cells from one tissue to another. Different studies have found that up- or down-regulation of -actinin could be related with different types of cancer, such as hepatocellular carcinoma (Nishiyama et al. 1990), • colorectal cancer (Honda et al. 2005), lung cancer (Honda et al. 2004), breast cancer (Guvakova et al. 2002) and neuroblastoma (Nikolopoulos et al. 2000).

4. Regulators of -actinin 4.1 Regulation by calcium In non-muscle cells the calcium-binding domain of -actinin contains EF-hands motifs that are able to bind calcium. Calcium concentrations less than 0.1 M decrease F-actin viscosity considerably (Blanchard et al. 1989). However, depending on the isoform the affinity for calcium is different. For instance, chicken -actinin-4 binds calcium with lower affinity than -actinin-1, suggesting different functions for these two isoforms (Imamura and Masaki 1992; Imamura et al. 1994). The mechanism by how calcium inhibits actin-binding is still unknown. The recently determined crystal structure of the ABD has suggested a possible regulatory mechanism for the actin-binding

 26  Molecular characterization and evolution of -actinin: from protozoa to vertebrates inhibition by calcium. The loop connecting the two CH domains is in close proximity with the neck region connecting the ABD and the first spectrin repeat. It is proposed that when calcium is bound to the calcium-binding domain of one monomer, it interacts with the neck between the ABD and the first spectrin repeat of the other monomer hindering the rearrangement of the two CH, therefore, inhibiting the binding to actin (Franzot et al. 2005).

4.2 Regulation by phosphoinositides The interaction site of phosphoinositides with -actinin has been located at the CH2, immediately after the ABS3 (Fukami et al. 1996). In cultured cells, phosphoinositide binding regulates the rate of -actinin association-dissociation with actin filaments and integrins (Fraley et al. 2005). In muscle cells binding of phosphatidylinositol 4,5-bisphosphate release the calcium-binding domain from the linker between ABD and SR1 allowing binding to titin (Young and Gautel 2000).

4.3 Regulation by phosphorylation -actinin can be phosphorylated by kinases such as focal adhesion kinase. The ABD is phosphorylated on tyrosine 12 reducing the affinity of -actinin for actin. This would result in a reduction of contact with the cytoskeleton facilitating the turnover of the adhesion sites (Izaguirre et al. 2001).

 27  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

5. Molecular evolution of -actinin

5.1 Aim of the Study Many studies have been focused on the evolution and origin of spectrin and dystrophin from a common -actinin ancestor. Both spectrin and dystrophin have been identified in several organism and compared in detail (Baines 2003; Neuman et al. 2001; Roberts and Bobrow 1998). However, very little is known about the evolution of -actinin and its ancestral isoforms. The study of how the different -actinin isoforms originated is therefore an important key to reconstruct and understand the origin and evolution of this important superfamily of proteins. My research has been focused on the evolution and structure of -actinins from ancestral organism such as protozoa and fungi. After a thorough database search we have discovered many different -actinin isoforms. Some of these isoforms are atypical since they do not share all the common features of modern -actinins. These atypical • isoforms have been characterized and compared with known -actinins to determine whether or not they are genuine -actinins. As a result, a comprehensive and phylogenetic analysis of -actinin isoforms has been accomplished giving a better understanding not only of the evolution of the -actinin but also of the spectrin superfamily of proteins.

5.2 Evolution of the -actinin family. Paper I

5.2.1 -actinin isoforms When initiating this work, the number of known -actinin sequences was limited. These isoforms contained the typical hallmarks of -actinins: an actin-binding domain, a calcium-binding domain and a rod domain with four spectrin repeats. Our database mining discovered -actinin isoforms that included a wide number of organisms such as protozoa, fungi and metazoan. Some of these isoforms appeared to be typical -actinins, whereas other isoforms from fungi and protozoa only shared some of the characteristic domains. We retrieved a total of 26 -actinin isoforms from several genome databases; 16 -actinin from vertebrates, 4 from invertebrates, 4 from

 28  Molecular characterization and evolution of -actinin: from protozoa to vertebrates protozoa and 2 from fungi. We observed that vertebrates usually contained 4 distinct isoforms, while invertebrates contained only one. By using different primers we also cloned and sequenced an -actinin gene from the sea squirt Ciona intestinalis. This organism is supposed to be one of the closest relatives to vertebrates; therefore the study of its -actinin should give hints of what happened in the evolution of -actinin before the vertebrate-invertebrate split. All the sequences were analyzed in silico to identify conserved -actinin domains. Vertebrate and invertebrates isoforms contained the typical hallmarks of -actinins including a rod domain with four spectrin repeats. When we analyzed the isoforms of Encephalitozoon cuniculi and the protozoans Trichomonas vaginalis and Entamoeba histolytica, we identified an N-terminal actin-binding domain and a C-terminal calcium-binding domain. However, the central rod domain of these isoforms was atypical; the protozoa isoforms contained a single repeat and that of fungi contained two spectrin repeats (Figure 13).

A BD S R 1 S R 2 CBD Fungi

CC A BD S R 1 CBD T richom onas vaginalis

A BD CC CBD E ntam oeba histolytica

Figure 13. Schematic drawing of the atypical alpha actinins. The short rod domain of these isoforms contains one or two putative spectrin repeats. ABD - actin-binding domain; SR - spectrin repeats; CBD - calcium-binding domain CC - coiled coil

5.2.2 Phylogenetic tree A phylogenetic analysis of all the -actinin isoforms gave a tree with three main clusters (figure 14): one with all vertebrate isoforms, a second with the invertebrates isoforms, and a third and distanced cluster corresponding to the protozoa and fungi. The phylogeny placed the atypical isoforms as the most divergent from vertebrate -actinins. These results were in agreement with our

 29  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

expectations, since we considered that these atypical isoforms should be the most ancestral isoforms and therefore the most divergent. The observation that the single isoform of Ciona intestinalis is placed right before vertebrates indicates that the four different -actinin sequences likely arose after the vertebrate radiation. In the vertebrate branches, the four isoforms are differently clustered forming two clear groups, one represented by the calcium-sensitive isoforms -actinin-1 and -actinin-4, and a second group formed by the calcium-insensitive -actinin-2 and -actinin-3. This topology is in accordance with the hypothesis that two rounds of genome duplication occurred early in the vertebrate divergence.

•

Figure 14. Phylogenetic analysis of the -actinin family. The most parsimonious tree shows the relation between -actinin of different organisms. Bootstrap values show the reliability of each branch.

5.2.3 Spectrin repeats We compared the vertebrate spectrin repeats (SR1-SR4) with the single repeat of E. histolytica and the two repeats of fungi. The E. histolytica repeat appears to be closest to SR1 of vertebrates whereas the two repeats in Schizosaccharomyces pombe were more similar to vertebrate SR1 and SR4 respectively. This led us to suggest a possible

 30  Molecular characterization and evolution of -actinin: from protozoa to vertebrates evolutionary scenario for the spectrin repeats in -actinin: the most ancestral -actinin contained a single repeat related with SR1 of modern -actinins, a first intragene duplication gave rise to a second repeat, closely related to SR4 in modern -actinins. A second subsequent intragene duplication gave rise to the central repeats SR2 and SR3 present in modern -actinins. The phylogeny of the rod domain in -actinins as well as a comparative and exhaustive analysis of the spectrin repeats will be discussed below in paper IV.

5.3 Isolation and characterization of atypical -actinins from Entamoeba histolytica. Paper II and III One of the most important issues for us was to determine whether these atypical isoforms were genuine -actinin or whether they simply shared some common domains. When the genome of the protozoa Entamoeba histolytica was completed we identified a second member of the -actinin family, this isoform contained two spectrin repeats and shared high homology with -actinin isoforms found in fungi. Entamoeba histolytica is a primitive unicellular organism that induces amoebiasis, one of the three most common human disease caused by parasites. It has been seen that reorganization of the actin cytoskeleton and actin-binding proteins plays an important role in the parasite infection mechanisms. We have therefore used Entamoeba histolytica as a model organism for the study of ancestral -actinin isoforms. We isolated and characterized these two -actinin isoforms and demonstrate that they are genuine -actinins.

5.3.1 Structure prediction

-actinin-1 (paper II) The molecular weight of -actinin1 is 63 kD. This isoform conserves the actin-binding domain with two CH domains and the calcium-binding domain where at least three EF-hands are recognized.

 31  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

The rod domain is shorter than typical -actinin and is predicted to form a coiled-coil. -actinin-2 (paper III) The molecular weight of -actinin-2 is 70 kD. Protein structure prediction and modelling detected a conserved actin-binding domain, a calcium-binding domain, and a rod domain with two spectrin repeats. This protein contains the same number of repeats as fungi -actinins. As observed in fungi, the two repeats present in Entamoeba histolytica -actinin-2 are closely related to SR1 and SR4 of typical -actinin, respectively.

5.3.2 Biochemical analyses We have examined among other features whether these two proteins are real -actinins. The main function of -actinin is to cross-link actin filaments; additionally, in some isoforms calcium can perturb the binding to actin. Both genes corresponding to -actinin-1 and -actinin-2 were cloned from an E. histolytica genome library. The corresponding proteins were expressed and biochemically characterized in terms of • calcium-binding and cross-linking ability.

Actin-binding properties As mentioned above there are several methods to study the actin-binding properties of a protein. The actin-binding properties of -actinin-1 and -actinin-2 were determined in presence and absence of calcium using two different methods: co-sedimentation assay and transmission electron microscopy. The co-sedimentation experiments showed that actin was gradually pelleted with increasing amounts of -actinin-1 or -actinin-2. This clearly indicated that both isoforms cross-linked actin. These results were also supported by the electron micrographs, where compact and tight bundles were observed in the presence of either -actinin-1 or -actinin-2. When the same experiments were repeated in the presence of calcium, both the amount of pelleted actin and the bundle formation were reduced. This indicated that calcium inhibits actin-binding to -actinin.

 32  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Calcium-binding properties The calcium-binding properties of both proteins were tested by a calcium overlay assay using 45Ca. The results showed clearly that both proteins bind calcium.

Dimer formation It is very likely that both isoforms form dimers since otherwise the cross-linking of actin filaments would not be possible. Both proteins were analyzed by gel filtration chromatography to study their native size. We observed that both -actinin-1 and -actinin-2 gave rise to gel filtration elution peaks that corresponded to monomeric and dimeric species.

5.3.3 Conclusions of paper II and paper III The two -actinin-like proteins present in E. histolytica conserve the essential domains of -actinins: the actin-binding domain and the calcium-binding domain. The rod domains of these two isoforms are shorter than the rod domain of typical -actinins. The rod domain of -actinin-1 contains a putative repeat predicted to form a coiled-coil structure whereas the rod domain of -actinin-2 contains, although with a low score, two predicted spectrin repeats. Despite of the shorter rod domain, both isoforms behave as modern -actinins. We conclude that both -actinin-1 and -actinin-2 are able to form dimers and cross-link actin in a calcium-dependent manner. Further, these proteins are detected by polyclonal chicken -actinin antibodies and share the high -helix content typical of -actinins. The existence of these proteins in an ancestral organism makes them good candidates to represent ancestral isoforms of -actinins and possible ancestors of this family.

 33  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

5.4 Evolution of -actinin rod domain. Paper IV Among all -actinin isoforms, the rod domain is the most heterogeneous domain concerning amino acid sequence and number of spectrin repeats. As I have mentioned in the introduction, all members of the spectrin superfamily are thought to have arisen by intragene duplications and other rearrangements from a common -actinin ancestor with four spectrin repeats. We propose that the origin of the spectrin superfamily started much earlier from a single repeat -actinin that experienced two rounds of intragene duplication to give rise to -actinin with four spectrin repeats. The finding of more atypical -actinins and the availability of other isoforms from ancestral organisms, have contributed to reinforce our previous hypothesis as well as to achieve other important understandings of the origin and evolution of this family. We have done a comprehensive and phylogenetic analysis of the rod domain studying all spectrin repeats present in most -actinin isoforms available to date. • 5.4.1 Pairwise identities The most conserved repeat in all organism is the SR1, probably due to its role in actin-binding as it was already seen in its close relative spectrin (Li and Bennett 1996). A general observation is that the differences between the spectrin repeats within the same isoforms are much larger than between the same repeat in -actinins of other organism. This is also observed in E. histolytica -actinin-2 and fungi repeats. As expected, the vertebrate spectrin repeats are the best conserved followed by the invertebrates. The least conserved repeats are those in protozoa and fungi. The typical conserved tryptophan residues at position 17 and close to the C-terminal are present in almost all vertebrate and invertebrate repeats. However there are exceptions like the two repeats in fungi and E. histolytica -actinin-2, Dictyostelium SR4 and Phytophthora SR3 and SR4. It is noteworthy that those repeats were still recognised as spectrin repeats by the SMART and InterPro servers.

 34  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

5.4.2 Phylogeny of the spectrin repeats A phylogenetic tree with the aligned sequences was constructed using Bayesian inference (figure 15). The most important feature of the tree is the presence of four clusters corresponding to the four spectrin repeats (SR1, SR2, SR3, and SR4). We could observe that SR1 and SR4 were the most separated branches. As already noticed in the pairwise identity analysis the different repeats in each -actinin isoform tend to group with the same repeat in isoforms of other organism rather than with themselves. Phytophthora isoforms constitutes an exception since SR2, SR3 and SR4 group in a single cluster. This indicates that they are more similar to each other than to the respective repeat in other organism isoforms. The two repeats present in fungi and E. histolytica -actinin-2 (SR1 and SR2) aligned with the SR1 and SR4 of modern -actinin respectively. The single repeat present in Trichomonas also grouped with SR1.

Figure 15. Phylogenetic tree of the spectrin repeats. The tree was calculated by Bayesian interference and the probabilities of each branch are indicated.

 35  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

5.4.3 Discussion and conclusions of paper IV -actinin isoforms can be divided in three main groups, one group contains a putative single repeat like -actinin-1 in E. histolytica, E. cuniculi or T. vaginalis. A second group contains two spectrin repeats like -actinin in fungi and -actinin-2 in E. histolytica. The third group comprises modern -actinins with four spectrin repeats present in all metazoan and other groups such as Dictyostelium or Phytophthora. Although with some exceptions we could say that the general trend is that the same repeats are more similar to each other than the other repeats in the isoform. The evolution of -actinin rod domain can be divided in two phases: A first phase of intragene duplications. This phase would start with an -actinin with a single repeat closely related to SR1 of modern organisms. A first intragene duplication gave rise to organism with two spectrin repeats where the new second repeat would be close to SR4 of modern -actinins. Finally a second intragene duplication added two more repeats that became SR2 and SR3. The fact that some of the repeats in Phytophthora -actinin are closer • to each other rather than to the same repeat in other isoforms, could suggest that in an early evolutionary stage these repeats still conserved similarities as a result of a duplication from a common ancestor. A second phase of consolidation In this phase the number of repeats was constant and the different repeats started to evolve independently to acquire specialized roles. It is quite clear that of all the spectrin repeats SR1 is the most conserved, regardless of the organism. We believe that this repeat is the most ancestral, and probably in early organisms did not contain all of the conserved residues present in a typical spectrin repeat. The conservation of this repeat during evolution is probably related with its role in actin-binding. We observed that SR1 and SR2 of fungi are quite divergent, and this is also observed between SR1 and SR4 of ancestral organism such as Dictyostelium and Phytophthora. In addition it has been seen that SR1 in fungi and modern isoforms are more basic than the other repeats. We suggest that after the first intragene duplication the resulting second repeat evolved to be more distinct and favour dimerization.

 36  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

6. Final Conclusions

All -actinins isoforms have conserved the ABD and the calcium-binding domain. However, the central rod domains are quite divergent among ancestral organism. -actinin isoforms can be divided in three different groups depending on the number of repeats present in the rod domain. One group with single-repeated -actinins, a second group represented by -actinins with two repeats, and a third group corresponds to modern -actinin with four repeats. We have shown that the two -actinin isoforms present in Entamoeba histolytica are genuine -actinins and therefore can be considered for phylogenetic and other comparative studies. There is also evidence that T. vaginalis and S. pombe -actinins are also genuine -actinins (Addis et al. 1998; Wu et al. 2001). Our phylogenetic studies placed these atypical -actinins at the bottom of the evolutionary tree of -actinin suggesting they are ancestral isoforms. The phylogeny also shows that is it not until the vertebrate radiation that the four -actinin genes appeared as result of two rounds of genome duplication from the single invertebrate -actinin gene. These four different genes gave rise to at least four different isoforms. In our phylogenetic tree the two calcium-sensitive isoforms (-actinin-1 and -actinin-4) cluster together in the same manner as the non-calcium sensitive isoforms (-actinin2- -actinin-3). Since evolution tends to eliminate redundant genes, the presence of two different calcium-sensitive isoforms and two calcium-insensitive isoforms suggests a high degree of specialization for these isoforms. In fact, each isoform can be found in different tissues having specialized functions. The diversity may be even more complicated since some genes can be alternatively spliced. Therefore it can be concluded that -actinin despite of being the shortest member of the spectrin superfamily has very specialised functions. We have suggested that the most ancestral -actinin contained a single spectrin-like repeat closed related to SR1. A first intragene duplication gave rise to isoforms with two repeats and a second intragene duplication then gave rise to isoforms with four repeats. This active phase would be dominated first by concerted evolution and intragene duplication. A second phase would come later where each of the four repeats evolved independently.

 37  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

It is therefore proposed that at this point -actinin evolution branched into different pathways, one would be the consolidation of the -actinin lineage, and other would be the elongation and duplication of the repeats giving rise to the different members of the spectrin superfamily.

•

 38  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

7. Future perspectives

Although we have achieved many important insights in the evolution of -actinin there are still other important issues to study. As the sequencing of different genomes is being completed, new -actinin isoforms are being identified. The most interesting isoforms from the evolutionary point of view are those present in ancestral organisms. It is therefore very important that these new sequences are added, analyzed and compared to strengthen our conclusions or suggest new ones. It would be of a great importance to purify and characterize other atypical isoforms as we have done with the two E. histolytica isoforms. Although structure predictions are already of a great accuracy, we would like to crystallize some of the atypical repeats and particularly the -actinin with one repeat present in Entamoeba histolytica to determine if the repeats fold in the same manner as common spectrin repeats.

 39  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Appendix:

Phylogenetic analysis Phylogenetic analysis has become a very powerful tool to understand the relationship among different proteins by comparing similarities or identities. Examination of the phylogeny of one protein family can determine how closely related one sequence is to a sequence which protein function is known. The most common and easy way to represent phylogeny is by creating phylogenetic trees.

Phylogenetic trees The first and most important step in order to build a reliable phylogenetic tree is to make an accurate alignment since the tree will be based on it. This is sometimes difficult since the sequences can be quite divergent. It is possible to manually refine the alignment, especially if there are known regions of structural similarities. A phylogenetic tree is a graphical representation of the relationship • among different proteins. A tree consists of two elements: nodes and branches. Branches are lines that intersect and terminate at nodes. At the tips of the branches we find the taxa or sequences. There are different styles to represent trees: a cladogram only displays the branching order whereas a phylogram displays both branching order and information about evolutionary rates through the lengths of the different branches. In the simplest case, branch lengths are proportional to the number of amino acid substitutions that has taken place along the branches. A tree can be rooted if there is a common ancestor to all other taxa in the analysis. An unrooted tree only shows the relations among the taxa without specify evolutionary pathways. There are several methods for calculating trees; none of them is worse or better, they are based on different assumptions and can be used in different circumstances. Usually, similar or identical topologies are obtained regardless of the method if the sequences are clearly related and conserved. The most used methods for protein inference are distance methods and character based methods. If the different methods produced different trees, one should look for regions that are similar regardless of the method used (Hall 2001).

 40  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Distance matrix methods Distances are expressed by the number of differences between two sequences in a certain alignment. They give only one possible tree. UPGMA (Unweighted Pair Gap Method with Arithmetic Mean) This method uses a sequential clustering algorithm. The building of the tree starts by pairing the two sequences with the smallest distances and combining them into a single cluster; the process is repeated for the remaining sequences. This method is very sensitive to the order of which the sequences are added. Neighbour-joining In this algorithm the distances between all the sequences are calculated and a distance matrix is constructed. The nodes are combined until a tree that gives the smallest sum of branch lengths is found. This method is not recommended for very divergent sequences. One of the inconveniences using this method is the long branch attraction effect, meaning that long branches representing rapidly evolved sequences tend to be aligned with other rapidly evolved sequences even though they might not be related at all. Since data is not used directly but rather the derived matrix, some data can be lost.

Character-based methods The character based methods use the sequence alignment directly rather than distance matrixes. Maximum likelihood. This method chooses the tree with the highest probability (likelihood) for the given data (the alignment) and a certain model. This model assumes that each site evolved independently and each site has certain likelihood. The probability of each tree is the product of the site probabilities. Parsimony. The tree is created by minimizing the changes needed to explain the data. With this method it is possible to obtain more than one equally parsimonious tree. The philosophy of parsimony is that nature will always choose the simplest scenario. Each site in the alignment is considered as a character; the inconvenience is that characters that are present in all data or only present in one taxon are ignored. Parsimony assumes that if two sequences have a common character it is because it was inherited from a common ancestor. Convergence (the taxa inherited the common characteristic independently) is not considered.

 41  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Bayesian method. This is a recent variant of maximum likelihood. It searches for the trees with the greatest likelihoods given the data. The difference with maximum likelihood is that Bayesian method produces several trees with high likelihood instead of a single one. This method does not use bootstrap values, instead the probability of a certain branch is given by the frequency of that branch in the set of trees. The most popular program for calculating trees by Bayesian inference is MrBayes. MrBayes starts with one tree (random or not), examines the tree according to the model, modifies the tree, examines the new tree again, and if the new tree is more probable it will accept it. This is “one generation”. In every generation the best tree will be recorded and used to create the consensus tree. MrBayes uses “posterior probabilities” that are calculated after knowing something about the data assuming a certain model (Huelsenbeck et al. 2001).

Tree reliability As mentioned before there is not a best method to construct phylogenetic trees. Every method has its own advantages and • disadvantages. To guarantee maximum reliability one should apply different methods and different settings, add or remove sequences to see the influence on the topology. It is also recommended to observe if the resulting tree for the studied protein is in agreement with trees obtained by using other sequences. A statistical evaluation can be also done using bootstrapping or jackknife methods. The bootstrapping method takes different subsample of the alignment and makes a tree. This is repeated several times (defined by the user and usually between 100-1000). The results will indicate the number of times that a specific branch is observed in the subsets. Jackknife removes one sequence, reanalyses the data several times and observes the frequency of a given branch (Hall 2001).

 42  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Acknowledgments

Here I am at last. This page is the most important for me, because without all the people I mention below, the previous pages would never have been written. First of all I would like to thank all the people who helped with my thesis before printing it. I want to thank the people at the Biochemistry Department for interesting seminars, courses, feedback and protocols. Thanks for being my little “Biochemistry family". Thanks to Anna Märta, for all the translation help with personal and administrative stuff. Lars: Thank you for showing me that big problems do not exist. Thank you for sometimes being a little tough at the lab (I promise you I will never say pH=7.6 is the same as pH=7.5 again, although it is almost the same, isn’t itJ). I love my work and I thank you for that! Thanks to the Spanish people around me for making me feel at home when I needed it. A part of me is missing when I am not talking Spanish. Thanks Maribel for our "phylogenetic" conversations. Thanks Miguel for the papers you “smuggled” for me, I have saved many hours searching in the library! Thanks to the guys, Jocke and Alvaro; I had a very nice time with you both. Alvaro, thanks for all your help in the lab and for the Flash presentation. You should have copyrighted it! Thanks to the former project students, Lisa, Mimi, Nina, was nice to share an office and some parties with you. Barbarita: thanks for all the funnnnnnnnn!! We have had together, inside and outside the lab. I will never have such a chic and fancy lab mate. Who will do my hair now...?? I wish you good luck with your research, you are very good and you deserve it! I will miss your “klick” “klock” in the corridor. I want to thank all the German community around me, thanks for making me feel German as well. Thanks Uta, for being the first German I met and sharing with me many good moments. Ruth, you’ve been my science advisor and my living-English dictionary (I will try to remember that “come along” is not the same as “come alone”!J) I admire your intelligence, naturalness and honesty (although the last two are not good for our poker eveningsJ).Thanks for cheering me up in my "I won’t make it" days, it meant a lot to me.

 43  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

We will still organize salsa evenings wherever we are in the world. It’s a Promise! Katja, you are my brute diamond, I admire your optimism and generosity. Thanks for the help in the lab. Thanks also for your true friendship and continuous affection to me. Remember you have a pending visit to the Trevi fountain…. Franziska, or “Francesca”?J You are the sweetest person I have meet, thanks for always being so nice to me. I hope you find that “unwrapped present”. I will continue sending you a post card wherever I am. I’ll miss our Wayne’s fikas! Anders, Thanks for standing by me when I was a little "jobbig". Thanks for being my taxi driver and coming to the lab with me very late in the evenings, at least you’ve learned that bacteria grow exponentially! J .This thesis is a prize I want to share with you. Now it is time for a little adventure…. Fasten your seat belt, we´re taking off…!! Thanks to all Anders friends and respective girlfriends for being always very "trevliga" to me, thanks for making me feel at home and showing me your country and culture. Thanks also for having the • patient to let me practice my crappie Swedish with you; otherwise it would have been much worse! Thanks to my Swedish girlfriends Emilia Jessica and Marie-Louise, you were the first friends I made in Umeå. Thanks for taking care of me! Victoria, gracias por buenas discusiones científicas y por la compañía en IKSU. Hecho de menos nuestras cenas en el colombiano (¿te conté que se cambió de restaurante?). Gracias por ponerme los pies en la tierra de vez en cuando. Te deseo un rápido camino al estrellato en lo laboral y personal ¡Te lo mereces! Gracias a mis amigas de España, por estar siempre allí desde que éramos pequeñitas. Por último, quiero agradecer a mi familia por siempre estar ahí a mi lado. A mis hermanos (Batiato e Ignony) a Sarita mi hermana, y a Papá y Mamá (Paco y Paca) gracias por vuestros consejos a lo largo de mi vida, Hoy no estaría aquí sin ellos.

 44  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

References

Addis MF, Rappelli P, Delogu G, Carta F, Cappuccinelli P, Fiori PL (1998) Cloning and molecular characterization of a cDNA clone coding for Trichomonas vaginalis alpha-actinin and intracellular localization of the protein Infect Immun, p 4924-31 Amann KJ, Renley BA, Ervasti JM (1998) A cluster of basic repeats in the dystrophin rod domain binds F-actin through an electrostatic interaction. J Biol Chem 273:28419-23 Atkinson RA, Joseph C, Kelly G, Muskett FW, Frenkiel TA, Nietlispach D, Pastore A (2001) Ca2+-independent binding of an EF-hand domain to a novel motif in the alpha-actinin-titin complex. Nat Struct Biol 8:853-7 Baines AJ (2003) Comprehensive analysis of all triple helical repeats in beta- spectrins reveals patterns of selective evolutionary conservation. Cell Mol Biol Lett 8:195-214 Bang ML, Mudry RE, McElhinny AS, Trombitas K, Geach AJ, Yamasaki R, Sorimachi H, Granzier H, Gregorio CC, Labeit S (2001) Myopalladin, a novel 145- kilodalton sarcomeric protein with multiple roles in Z-disc and I-band protein assemblies. J Cell Biol 153:413-27 Banuelos S, Saraste M, Carugo KD (1998) Structural comparisons of calponin homology domains: implications for actin binding. Structure 6:1419-31 Barstead RJ, Kleiman L, Waterston RH (1991) Cloning, sequencing, and mapping of an alpha-actinin gene from the nematode Caenorhabditis elegans. Cell Motil Cytoskeleton 20:69-78 Beggs AH, Byers TJ, Knoll JH, Boyce FM, Bruns GA, Kunkel LM (1992) Cloning and characterization of two human skeletal muscle alpha-actinin genes located on chromosomes 1 and 11. J Biol Chem 267:9281-8 Belkin AM, Koteliansky VE (1987) Interaction of iodinated vinculin, metavinculin and alpha-actinin with cytoskeletal proteins. FEBS Lett 220:291-4 Bennett H, Condeelis J (1988) Isolation of an immunoreactive analogue of brain fodrin that is associated with the cell cortex of Dictyostelium amoebae. Cell Motil Cytoskeleton 11:303-17 Bennett V, Baines AJ (2001) Spectrin and ankyrin-based pathways: metazoan inventions for integrating cells into tissues. Physiol Rev 81:1353-92 Blanchard A, Ohanian V, Critchley D (1989) The structure and function of alpha- actinin. J Muscle Res Cell Motil 10:280-9 Borrego-Diaz E, Kerff F, Lee SH, Ferron F, Li Y, Dominguez R (2006) Crystal structure of the actin-binding domain of alpha-actinin 1: evaluating two competing actin-binding models. J Struct Biol 155:230-8

 45  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Bricheux G, Coffe G, Pradel N, Brugerolle G (1998) Evidence for an uncommon alpha-actinin protein in Trichomonas vaginalis. Mol Biochem Parasitol 95:241-9 Broderick MJ, Winder SJ (2005) Spectrin, alpha-actinin, and dystrophin. Adv Protein Chem 70:203-46 Burridge K, Wennerberg K (2004) Rho and Rac take center stage. Cell 116:167-79 Byers TJ, Husain-Chishti A, Dubreuil RR, Branton D, Goldstein LS (1989) Sequence similarity of the amino-terminal domain of Drosophila beta spectrin to alpha actinin and dystrophin. J Cell Biol 109:1633-41 Castresana J, Saraste M (1995) Does Vav bind to F-actin through a CH domain? FEBS Lett 374:149-51 Cooper JA, Walker SB, Pollard TD (1983) Pyrene actin: documentation of the validity of a sensitive assay for actin polymerization. J Muscle Res Cell Motil 4:253- 62 Coulombe PA, Wong P (2004) Cytoplasmic intermediate filaments revealed as dynamic and multipurpose scaffolds. Nat Cell Biol 6:699-706 Davison MD, Baron MD, Critchley DR, Wootton JC (1989) Structural analysis of homologous repeated domains in alpha-actinin and spectrin. Int J Biol Macromol 11:81-90 • Disanza A, Steffen A, Hertzog M, Frittoli E, Rottner K, Scita G (2005) Actin polymerization machinery: the finish line of signaling networks, the starting point of cellular movement. Cell Mol Life Sci 62:955-70 Djinovic-Carugo K, Gautel M, Ylanne J, Young P (2002) The spectrin repeat: a structural platform for cytoskeletal protein assemblies. FEBS Lett 513:119-23 Djinovic-Carugo K, Young P, Gautel M, Saraste M (1999) Structure of the alpha- actinin rod: molecular basis for cross-linking of actin filaments. Cell 98:537-46 dos Remedios CG, Chhabra D, Kekic M, Dedova IV, Tsubakihara M, Berry DA, Nosworthy NJ (2003) Actin binding proteins: regulation of cytoskeletal microfilaments. Physiol Rev 83:433-73 Dubreuil RR (1991) Structure and evolution of the actin crosslinking proteins. Bioessays 13:219-26 Dubreuil RR, Byers TJ, Sillman AL, Bar-Zvi D, Goldstein LS, Branton D (1989) The complete sequence of Drosophila alpha-spectrin: conservation of structural domains between alpha-spectrins and alpha-actinin. J Cell Biol 109:2197-205 Ebashi S, Ebashi F (1965) Alpha-actinin, a new structural protein from striated muscle. I. Preparation and action on actomyosinatp interaction. J Biochem (Tokyo) 58:7-12

 46  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Endo T, Masaki T (1982) Molecular properties and functions in vitro of chicken smooth-muscle alpha-actinin in comparison with those of striated-muscle alpha- actinins. J Biochem (Tokyo) 92:1457-68 Ervasti JM, Campbell KP (1993) A role for the dystrophin-glycoprotein complex as a transmembrane linker between laminin and actin. J Cell Biol 122:809-23 Estes JE, Selden LA, Kinosian HJ, Gershman LC (1992) Tightly-bound divalent cation of actin. J Muscle Res Cell Motil 13:272-84 Fay FS, Fujiwara K, Rees DD, Fogarty KE (1983) Distribution of alpha-actinin in single isolated smooth muscle cells. J Cell Biol 96:783-95 Flood G, Rowe AJ, Critchley DR, Gratzer WB (1997) Further analysis of the role of spectrin repeat motifs in alpha-actinin dimer formation. Eur Biophys J 25:431-5 Fraley TS, Pereira CB, Tran TC, Singleton C, Greenwood JA (2005) Phosphoinositide binding regulates alpha-actinin dynamics: mechanism for modulating cytoskeletal remodeling. J Biol Chem 280:15479-82 Franzot G, Sjoblom B, Gautel M, Djinovic Carugo K (2005) The crystal structure of the actin binding domain from alpha-actinin in its closed conformation: structural insight into phospholipid regulation of alpha-actinin. J Mol Biol 348:151-65 Fujiwara K, Porter ME, Pollard TD (1978) Alpha-actinin localization in the cleavage furrow during cytokinesis. J Cell Biol 79:268-75 Fukami K, Sawada N, Endo T, Takenawa T (1996) Identification of a phosphatidylinositol 4,5-bisphosphate-binding site in chicken skeletal muscle alpha- actinin. J Biol Chem 271:2646-50 Gilmore AP, Parr T, Patel B, Gratzer WB, Critchley DR (1994) Analysis of the phasing of four spectrin-like repeats in alpha-actinin. Eur J Biochem 225:235-42 Gimona M, Djinovic-Carugo K, Kranewitter WJ, Winder SJ (2002) Functional plasticity of CH domains. FEBS Lett 513:98-106 Gluck U, Ben-Ze'ev A (1994) Modulation of alpha-actinin levels affects cell motility and confers tumorigenicity on 3T3 cells. J Cell Sci 107 ( Pt 7):1773-82 Guvakova MA, Adams JC, Boettiger D (2002) Functional role of alpha-actinin, PI 3- kinase and MEK1/2 in insulin-like growth factor I receptor kinase regulated motility of human breast carcinoma cells. J Cell Sci 115:4149-65 Hall B (2001) Phylogenetic Trees Made Easy: A How-To Manual for Molecular Biologist. Sinauer Associates, Inc. Honda K, Yamada T, Endo R, Ino Y, Gotoh M, Tsuda H, Yamada Y, Chiba H, Hirohashi S (1998) Actinin-4, a novel actin-bundling protein associated with cell motility and cancer invasion. J Cell Biol 140:1383-93

 47  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Honda K, Yamada T, Hayashida Y, Idogawa M, Sato S, Hasegawa F, Ino Y, Ono M, Hirohashi S (2005) Actinin-4 increases cell motility and promotes lymph node metastasis of colorectal cancer. Gastroenterology 128:51-62 Honda K, Yamada T, Seike M, Hayashida Y, Idogawa M, Kondo T, Ino Y, Hirohashi S (2004) Alternative splice variant of actinin-4 in small cell lung cancer. Oncogene 23:5257-62 Huelsenbeck JP, Ronquist F, Nielsen R, Bollback JP (2001) Bayesian inference of phylogeny and its impact on evolutionary biology. Science 294:2310-4 Imamura M, Masaki T (1992) A novel nonmuscle alpha-actinin. Purification and characterization of chicken lung alpha-actinin. J Biol Chem 267:25927-33 Imamura M, Sakurai T, Ogawa Y, Ishikawa T, Goto K, Masaki T (1994) Molecular cloning of low-Ca(2+)-sensitive-type non-muscle alpha-actinin. Eur J Biochem 223:395-401 Izaguirre G, Aguirre L, Hu YP, Lee HY, Schlaepfer DD, Aneskievich BJ, Haimovich B (2001) The cytoskeletal/non-muscle isoform of alpha-actinin is phosphorylated on its actin-binding domain by the focal adhesion kinase. J Biol Chem 276:28676-85 Katoh K, Kano Y, Masuda M, Onishi H, Fujiwara K (1998) Isolation and contraction • of the stress fiber. Mol Biol Cell 9:1919-38 Kliche W, Fujita-Becker S, Kollmar M, Manstein DJ, Kull FJ (2001) Structure of a genetically engineered molecular motor. Embo J 20:40-6 Knudsen KA, Soler AP, Johnson KR, Wheelock MJ (1995) Interaction of alpha- actinin with the cadherin/catenin cell-cell adhesion complex via alpha-catenin. J Cell Biol 130:67-77 Koenig M, Hoffman EP, Bertelson CJ, Monaco AP, Feener C, Kunkel LM (1987) Complete cloning of the Duchenne muscular dystrophy (DMD) cDNA and preliminary genomic organization of the DMD gene in normal and affected individuals. Cell 50:509-17 Koenig M, Monaco AP, Kunkel LM (1988) The complete sequence of dystrophin predicts a rod-shaped cytoskeletal protein. Cell 53:219-26 Kos CH, Le TC, Sinha S, Henderson JM, Kim SH, Sugimoto H, Kalluri R, Gerszten RE, Pollak MR (2003) Mice deficient in alpha-actinin-4 have severe glomerular disease. J Clin Invest 111:1683-90 Kremerskothen J, Teber I, Wendholt D, Liedtke T, Bockers TM, Barnekow A (2002) Brain-specific splicing of alpha-actinin 1 (ACTN1) mRNA. Biochem Biophys Res Commun 295:678-81 Kuhlman PA, Hemmings L, Critchley DR (1992) The identification and characterisation of an actin-binding site in alpha-actinin by mutagenesis. FEBS Lett 304:201-6

 48  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Lazarides E, Burridge K (1975) Alpha-actinin: immunofluorescent localization of a muscle structural protein in nonmuscle cells. Cell 6:289-98 Lazarides E, Lindberg U (1974) Actin is the naturally occurring inhibitor of deoxyribonuclease I. Proc Natl Acad Sci U S A 71:4742-6 Lemanski LF, Fuldner RA, Paulson DJ (1980) Immunofluorescence studies for myosin, alpha-actinin and tropomyosin in developing hearts of normal and cardiac lethal mutant Mexican axolotls, Ambystoma mexicanum. J Embryol Exp Morphol 55:1-15 Li X, Bennett V (1996) Identification of the spectrin subunit and domains required for formation of spectrin/adducin/actin complexes. J Biol Chem 271:15695-702 Liu J, Taylor DW, Taylor KA (2004) A 3-D reconstruction of smooth muscle alpha- actinin by CryoEm reveals two different conformations at the actin-binding region. J Mol Biol 338:115-25 Liu SC, Derick LH, Palek J (1987) Visualization of the hexagonal lattice in the erythrocyte membrane skeleton. J Cell Biol 104:527-36 Lundberg S, Lehto VP, Backman L (1992) Characterization of calcium binding to spectrins. Biochemistry 31:5665-71 Luther PK (1991) Three-dimensional reconstruction of a simple Z-band in fish muscle. J Cell Biol 113:1043-55 MacArthur DG, North KN (2004) A gene for speed? The evolution and function of alpha-actinin-3. Bioessays 26:786-95 Maruyama K, Ebashi S (1965) Alpha-actinin, a new structural protein from striated muscle. II. Action on actin. J Biochem (Tokyo) 58:13-9 McGough A (1998) F-actin-binding proteins. Curr Opin Struct Biol 8:166-76 McGough A, Way M, DeRosier D (1994) Determination of the alpha-actinin-binding site on actin filaments by cryoelectron microscopy and image analysis. J Cell Biol 126:433-43 Meyer RK, Aebi U (1990) Bundling of actin filaments by alpha-actinin depends on its molecular length. J Cell Biol 110:2013-24 Mills M, Yang N, Weinberger R, Vander Woude DL, Beggs AH, Easteal S, North K (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 10:1335-46 Neuman S, Kaban A, Volk T, Yaffe D, Nudel U (2001) The dystrophin / utrophin homologues in Drosophila and in sea urchin. Gene 263:17-29 Nikolopoulos SN, Spengler BA, Kisselbach K, Evans AE, Biedler JL, Ross RA (2000) The human non-muscle alpha-actinin protein encoded by the ACTN4 gene suppresses tumorigenicity of human neuroblastoma cells. Oncogene 19:380-6

 49  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Nishiyama M, Ozturk M, Frohlich M, Mafune K, Steele G, Jr., Wands JR (1990) Expression of human alpha-actinin in human hepatocellular carcinoma. Cancer Res 50:6291-4 Noegel A, Witke W, Schleicher M (1987) Calcium-sensitive non-muscle alpha- actinin contains EF-hand structures and highly conserved regions. FEBS Lett 221:391-6 Otey CA, Carpen O (2004) Alpha-actinin revisited: a fresh look at an old player. Cell Motil Cytoskeleton 58:104-11 Otey CA, Pavalko FM, Burridge K (1990) An interaction between alpha-actinin and the beta 1 integrin subunit in vitro. J Cell Biol 111:721-9 Otterbein LR, Graceffa P, Dominguez R (2001) The crystal structure of uncomplexed actin in the ADP state. Science 293:708-11 Parr T, Waites GT, Patel B, Millake DB, Critchley DR (1992) A chick skeletal- muscle alpha-actinin gene gives rise to two alternatively spliced isoforms which differ in the EF-hand Ca(2+)-binding domain. Eur J Biochem 210:801-9 Parry DA, Dixon TW, Cohen C (1992) Analysis of the three-alpha-helix motif in the spectrin superfamily of proteins. Biophys J 61:858-67 Pascual J, Castresana J, Saraste M (1997) Evolution of the spectrin repeat. Bioessays • 19:811-7 Perelroizen I, Didry D, Christensen H, Chua NH, Carlier MF (1996) Role of nucleotide exchange and hydrolysis in the function of profilin in action assembly. J Biol Chem 271:12302-9 Pollard TD (1984) Purification of a high molecular weight actin filament gelation protein from Acanthamoeba that shares antigenic determinants with vertebrate spectrins. J Cell Biol 99:1970-80 Roberts RG, Bobrow M (1998) Dystrophins in vertebrates and invertebrates. Hum Mol Genet 7:589-95 Roulier EM, Fyrberg C, Fyrberg E (1992) Perturbations of Drosophila alpha-actinin cause muscle paralysis, weakness, and atrophy but do not confer obvious nonmuscle phenotypes. J Cell Biol 116:911-22 Salmikangas P, Mykkanen OM, Gronholm M, Heiska L, Kere J, Carpen O (1999) Myotilin, a novel sarcomeric protein with two Ig-like domains, is encoded by a candidate gene for limb-girdle muscular dystrophy. Hum Mol Genet 8:1329-36 Small JV, Stradal T, Vignal E, Rottner K (2002) The lamellipodium: where motility begins. Trends Cell Biol 12:112-20 Speicher DW, Marchesi VT (1984) Erythrocyte spectrin is comprised of many homologous triple helical segments. Nature 311:177-80

 50  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Spudich JA, Watt S (1971) The regulation of rabbit skeletal muscle contraction. I. Biochemical studies of the interaction of the tropomyosin-troponin complex with actin and the proteolytic fragments of myosin. J Biol Chem 246:4866-71 Suzuki A, Goll DE, Singh I, Allen RE, Robson RM, Stromer MH (1976) Some properties of purified skeletal muscle alpha-actinin. J Biol Chem 251:6860-70 Tang J, Taylor DW, Taylor KA (2001) The three-dimensional structure of alpha- actinin obtained by cryoelectron microscopy suggests a model for Ca(2+)-dependent actin binding. J Mol Biol 310:845-58 Taylor KA, Taylor DW, Schachat F (2000) Isoforms of alpha-actinin from cardiac, smooth, and skeletal muscle form polar arrays of actin filaments. J Cell Biol 149:635-46 Thomas GH, Newbern EC, Korte CC, Bales MA, Muse SV, Clark AG, Kiehart DP (1997) Intragenic duplication and divergence in the spectrin superfamily of proteins. Mol Biol Evol 14:1285-95 Trave G, Pastore A, Hyvonen M, Saraste M (1995) The C-terminal domain of alpha- spectrin is structurally related to calmodulin. Eur J Biochem 227:35-42 van den Ent F, Amos LA, Lowe J (2001) Prokaryotic origin of the actin cytoskeleton. Nature 413:39-44 Wade RH, Meurer-Grob P, Metoz F, Arnal I (1998) Organisation and structure of microtubules and microtubule-motor protein complexes. Eur Biophys J 27:446-54 Waites GT, Graham IR, Jackson P, Millake DB, Patel B, Blanchard AD, Weller PA, Eperon IC, Critchley DR (1992) Mutually exclusive splicing of calcium-binding domain exons in chick alpha-actinin. J Biol Chem 267:6263-71 Walikonis RS, Oguni A, Khorosheva EM, Jeng CJ, Asuncion FJ, Kennedy MB (2001) Densin-180 forms a ternary complex with the (alpha)-subunit of Ca2+/calmodulin-dependent protein kinase II and (alpha)-actinin. J Neurosci 21:423- 33 Wasenius VM, Narvanen O, Lehto VP, Saraste M (1987) Alpha-actinin and spectrin have common structural domains. FEBS Lett 221:73-6 Way M, Pope B, Weeds AG (1992) Evidence for functional homology in the F-actin binding domains of gelsolin and alpha-actinin: implications for the requirements of severing and capping. J Cell Biol 119:835-42 Winder SJ, Ayscough KR (2005) Actin-binding proteins. J Cell Sci 118:651-4 Winder SJ, Gibson TJ, Kendrick-Jones J (1995a) Dystrophin and utrophin: the missing links! FEBS Lett 369:27-33 Winder SJ, Hemmings L, Maciver SK, Bolton SJ, Tinsley JM, Davies KE, Critchley DR, Kendrick-Jones J (1995b) Utrophin actin binding domain: analysis of actin binding and cellular targeting. J Cell Sci 108 ( Pt 1):63-71

 51  Molecular characterization and evolution of -actinin: from protozoa to vertebrates

Winkler J, Lunsdorf H, Jockusch BM (1997) Flexibility and fine structure of smooth-muscle alpha-actinin. Eur J Biochem 248:193-9 Witke W, Hofmann A, Koppel B, Schleicher M, Noegel AA (1993) The Ca(2+)- binding domains in non-muscle type alpha-actinin: biochemical and genetic analysis. J Cell Biol 121:599-606 Wu JQ, Bahler J, Pringle JR (2001) Roles of a fimbrin and an alpha-actinin-like protein in fission yeast cell polarization and cytokinesis. Mol Biol Cell 12:1061-77 Wyszynski M, Lin J, Rao A, Nigh E, Beggs AH, Craig AM, Sheng M (1997) Competitive binding of alpha-actinin and calmodulin to the NMDA receptor. Nature 385:439-42 Ylanne J, Scheffzek K, Young P, Saraste M (2001) Crystal structure of the alpha- actinin rod reveals an extensive torsional twist. Structure 9:597-604 Young P, Ferguson C, Banuelos S, Gautel M (1998) Molecular structure of the sarcomeric Z-disk: two types of titin interactions lead to an asymmetrical sorting of alpha-actinin. Embo J 17:1614-24 Young P, Gautel M (2000) The interaction of titin and alpha-actinin is controlled by a phospholipid-regulated intramolecular pseudoligand mechanism. Embo J 19:6331- 40 • Zaner KS, Stossel TP (1982) Some perspectives on the viscosity of actin filaments. J Cell Biol 93:987-91

 52 