cells work is fundamental to all biological systems. It is very important to appreciate

Transcription

1 ! CHAPTER 1: INTRODUCTION 1.1 CELL BIOLOGY AND THE CYTOSKELETON Cells are the basic unit of life, which is what makes cell biology one of the most important areas of research. Though cells have been studied for several centuries, new discoveries are still being made that pertain to cellular organelles structures and signaling molecules. Knowing the components of the cell and understanding how cells work is fundamental to all biological systems. It is very important to appreciate the similarities and differences between cell types not only for a better understanding but also to advance the biomedical fields such as developmental biology and cancer research. Many aspects of the cell have been studied in great detail and the cytoskeleton is one of them. It is a vital component of all cells whether prokaryotic or eukaryotic. Nikolai Konstantinovich Koltsov ( ), a Russian biologist in the year 1903, was the first to propose that cells had an internal support network, that later became known as the cytoskeleton. The cytoskeleton is many things to the cell: a structural scaffold defining the cell shape, an intracellular transport system, a driver of cell motility, and a mediator of cell division, to name a few of the most important. In other words, just as skeleton and muscle function to provide structure for the human body, the cytoskeleton gives structure for the cell. Eukaryotic cytoskeleton: The cytoskeleton in eukaryotic cells has been studied extensively. It is composed of a framework of at least three classes of fibres that include proteins. It comprises microfilaments, intermediate filaments and microtubules together with their respective associated proteins (Steinmetz et al., 1997).! $

2 Polymers of tubulin, termed Microtubules, extend throughout the cell, providing an organizational framework for organelles in the cell. Microtubules are about 25nm in diameter (Nogales et al., 1999). The primary role of the microtubule cytoskeleton is mechanical support, although microtubules also take part in many other processes. They also form the spindles that pull sister chromatids apart during mitosis. Microtubules also give shape to cilia and flagella in eukaryotes. Actin polymers are known as Microfilaments and are known to have a diameter of 6 nm (Fuchs and Cleveland, 1998). They can form complex structures by being held together in different ways by a number of actin-binding proteins. In a typical eukaryotic cell (animal, plant, or a eukaryotic microorganism), complexes of cytoskeletal elements are located in a layer close to the inside of the cytoplasmic membrane and found to interact with it. This layer, just under the membrane known as cell cortex, is an actin-rich layer in animal cells. Microfilaments interact with and play a role in organizing the plasma membrane. Besides forming the contractile ring that divides the cell into two during mitosis (Mayer, 2005), actin is also seen as a fibrous system across the cytoplasm. Intermediate filaments are composed of a variety of proteins unlike actin filaments and microtubules that are polymers of single types of proteins. They have a diameter of about 10 nm, which is intermediate between the diameters of the other two elements of the cytoskeleton (Fuchs and Cleveland, 1998). Intermediate filaments are tissue-specific and are expressed in different types of cells. They are known to play a structural role by providing mechanical strength to cells and tissues (Toivola et al., 2005). Keratin genes account for most of the intermediate filament genes in the human genome (Schweizer et al., 2006).! 5

3 Intermediate filaments are not known to have a transport role i.e. they are not substrates for motor proteins to walk upon. By contrast, microfilaments and microtubules both serve as transport systems. Eukaryotic cells use motor proteins that use cytoskeletal polymers as tracks to transport various intracellular cargos, including membranous organelles, protein complexes and mrnas (Vale, 2003). Three classes of cytoskeletal motors are known: myosin, which interacts with actin filaments for transport of cargo, and two types of microtubule motors, dynein and kinesin. While conventional kinesins, and most other kinesins, move towards the plus end, one kinesin family is known to move bidirectionally. Dyneins on the other hand, move towards the microtubule minus ends. Molecular motors convert the chemical energy derived from ATP hydrolysis into mechanical action, and precise spatial and temporal regulation of motor-based transport ensures efficient cargo delivery (Woehlke and Schliwa, 2000). Data from sequencing the human genome revealed that more than 800 genes code for components of the cytoskeleton or proteins interacting with the cytoskeleton. These findings underlie the importance of the cytoskeleton for the cell. Prokaryotic cytoskeleton: In contrast to the situation in eukaryotic cells, the presence of a well-defined cytoskeleton in prokaryotes remained unknown for decades (Mayer, 2005); (Nanninga, 2001). The existence of a true bacterial cytoskeleton became apparent when the sugar kinases and prokaryotic cell cycle proteins were found to resemble actin structurally (Bork et al., 1992). It soon became evident that the FtsZ polypeptide had a similar macromolecular architecture as that of tubulin- the structural element of microtubules in eukaryotes (Erickson et al., 1996; Lowe and Amos, 1998). The bacterial protein MreB, thought to belong to the actin superfamily, was discovered! 9

4 only about a decade ago by Jones et al. (Jones et al., 2001). With advances in visualization technology and structure determination of the cells, analogues for all major cytoskeletal proteins in eukaryotes were found in prokaryotes. Furthermore, cytoskeletal elements with no known eukaryotic homologue have also been discovered in prokaryotes (Wickstead and Gull, 2011). Even in prokaryotes, cytoskeletal elements play vital roles in cell division (FtsZ), shape determination (Crescentin in Caulobacter crescentus) and polarity determination (MreB) (Michie and Lowe, 2006; Shih and Rothfield, 2006). 1.2 ACTIN CYTOSKELETON Actin is a highly conserved ATPase that assembles into cytoskeletal polymers called microfilaments. It is a eukaryotic protein that is expressed in most animals, plants and fungi in many isoforms (Hennessey et al., 1993). Actin has been extensively studied over several decades despite which, there still exists a great interest in actin probably due to its dynamic involvement in many aspects of cell biology. W.D. Halliburton, who extracted a protein from muscle that coagulated preparations of myosin, first observed actin experimentally in 1887 and he called it "myosin-ferment" (Halliburton, 1887). However, Halliburton was unable to further refine his findings, and the discovery of actin is credited instead to Brunó Ferenc Straub. In 1942, Straub developed a novel technique for extracting muscle protein that allowed him to isolate substantial amounts of relatively pure actin. Straub's method is essentially the same as that used in laboratories today. Hanson & Lowy first visualized actin filaments in 1963 by electron microscopy in one of the pioneering applications of negative staining techniques (Hanson, 1963). The filaments appeared as two strings of beads! 6

5 that gradually twisted around each other with each bead being an actin monomer (Hanson, 1963) The 3-Dimensional Structure of Actin Actin is a globular protein (G-actin) with a molecular weight of!42 kda, which polymerizes into filaments (F-actin) for most of its biological functions (Carlier, 1991; Korn et al., 1987). Actin exists in a dynamic equilibrium between these two states of monomer and filamentous polymer. Atomic structure of actin monomer Some of the most significant developments in the current understanding of actin come from the determination of its atomic structure from crystals of actin. Because conditions favoring crystallization of G- actin also induce its polymerization into F- actin filaments, it has been difficult to obtain 3-D crystals of pure actin suitable for X- ray diffraction analysis. To overcome this problem, several investigators have cocrystallized actin with proteins that inhibit its polymerization thereby making it suitable for X-ray analysis. These efforts made it possible to elucidate the atomic structures of actin in complex with G-actin binding proteins such as DNase I (Kabsch et al., 1990), Profilin (Schutt et al., 1989), Gelsolin fragment 1 (Mannherz et al., 1992) and the muscle protein, Myosin. These studies have provided tremendous amounts of valuable information (Amos et al., 1982; Elzinga et al., 1973). Actin is a disc like polypeptide made up of ~375 amino acid residues. The actin monomer has two major domains: a small domain (1) that comprises residues and and a large domain (2) that comprises residues Each of these domains can be further divided into two subdomains. Subdomains I (residues 1-32, , and ) and II (residues 33-69) form the two subdomains of domain 1,! F

6 while subdomains III (residues and ) and IV (residues ) form domain 2 (Holmes, 1991) (See Illustration 1).!!!!"!!! Illustration 1: Actin monomer. A ribbon diagram of the four structural subdomains of actin shown in different colours. Subdomain I in red (residues 1-32, and ); Subdomain II in yellow (residues 33-69); Subdomain III in blue (residues and ); Subdomain IV in green (residues ). Figure modified from (Hennessey et al., 1993) The domains of the actin molecule are effectively separated by two diametrically opposed clefts, the target-binding cleft and the nucleotide-binding cleft (Dominguez, 2004). The target-binding cleft, also known as the hydrophobic cleft, laying between Subdomains I and III, is thought to mediate the interactions of actin with most ABPs (Dominguez, 2004). Each actin monomer binds an adenine nucleotide and its associated divalent cation in the nucleotide-binding cleft between Subdomains II and IV. The divalent cation is believed to be Mg 2+ under physiological conditions. The purine of ATP is sandwiched in a hydrophobic pocket between subdomains III and IV while the polyphosphate tail is held by two loops originating from subdomains I (P1- loop) and III (P2-loop) (Hennessey et al., 1993). The actin monomer can exist in different conformations depending on the status of the actin-bound ATP, the nature of the divalent cation either at its high-affinity site (Ca 2+ or Mg 2+ ) or at additional sites and the degree of oligomerization (Moraczewska et al., 1999; Schuler, 2001). Conformational changes of actin must occur in order to allow the ATP/ADP exchange to take place. From studies of the "-actin structure, the rotation of subdomain I of! M

7 actin with respect to subdomains III and IV is thought to be intrinsic to the opening of the nucleotide binding cleft (Schuler et al., 2006). Structure of F-actin A chain of monomers come together to form a helical filament of actin in which subunits are connected by a 167 rotation and 2.7 nm axial rise. Two actin protofilaments coil around each other to form a helical filament (Aebi et al., 1986). Assembly of actin monomers into a filament involves an initial nucleation step, which is inherently slow due to instability of actin dimers. However, once it assembles into a trimer, which is a more stable structure, the filaments can undergo rapid polarized growth. There is an actin monomer concentration threshold below which actin will not polymerize. This value is termed the Critical Concentration (Cc). In other words, the critical concentration is a measure of the ability of a solution of G-actin to polymerize, which is about 0.1 µm under in vitro conditions. At monomer concentrations above the Cc, actin filaments assemble until the free monomer concentration equals the Cc. Below the Cc value, a solution of F-actin will depolymerize (Lodish, 2000) The actin monomer must bind either ADP or ATP in order to polymerize. The Cc for assembly of actin filaments depends on whether the monomers are bound to ATP or ADP. An ATP-bound monomer has a much higher inclination towards polymerization than when ADP-bound. The monomeric ATP-actin subunits preferentially add to the barbed (+) end at a faster rate (11.6 subunits µm -1 sec -1 ) compared to that at the pointed (#) end (1.3 subunits µm -1 sec -1 ) (Pollard et al., 2000). This difference in elongation rates at the opposite ends of an actin filament is caused by the difference in Cc values at the two ends. The Cc is much lower for G-actin addition at the barbed end (Cc + = 0.1 µm) than that for addition at the pointed end (Cc - = 0.8 µm).! P

8 After the addition of ATP-subunit to the barbed end of a filament, the ATP bound to the subunit is hydrolyzed producing actin bound to ADP and Pi. In a subsequent (slower) step, the Pi is released from the subunit to leave an ADP bound actin subunit behind (Pollard et al., 2000). Depolymerization of actin filaments from the pointed end releases ADP-monomers back into the cytosol. These rapid events of actin filament turnover produce significant amounts of ADP actin monomers. Although theoretically these ADPbound actin monomers can polymerize too, the critical concentration required is 5-10 times higher than that for ATP-actin (Pollard, 1986). Since ATP-actin monomers are better suited for filament elongation, the cells recharge these ADP-monomers to ATPbound monomer. This process known as Nucleotide Exchange enables the monomers to be recycled for new rounds of polymerization. The exchange of nucleotide bound to the actin is a potential regulatory step in the assembly of the cytoskeleton (Carlier, 1989; Pollard, 1986). The extent of polymerization of the bulk of actin monomers and the stability of an existing actin filament depends on whether ADP or ATP is bound to the actin monomers. The concentrations of the ion bound to the actins usually (most often Mg2+, sometimes Ca2+, K+ or Na+) determine whether filament formation or dissociation is favoured. Eventually the filament will grow to where the ambient concentrations of G-actin and requisite ions have been depleted enough that the rate of actin monomer dissociation from the filament equals the rate of elongation achieving a Steady-state. Filaments at steady state, undergoing addition and loss at their ends, display a Treadmilling behavior. The length of the filament thus remains constant, with the newly added subunits traveling through the filament as if on a treadmill until they reach the (#) end where they dissociate. These filaments are thus a mosaic enriched in ATP bound actin! R

9 near the barbed (+) end, ADP-Pi actin in the middle and ADP-actin at the pointed (#) end (Nicholson-Dykstra et al., 2005) (See Illustration 2). All actin monomers in the filament orient with their cleft toward the same end of the filament (designated the minus end) assigning a polarity to the actin filament. The two larger subdomains, Subdomains I and III, that form a flexible base of the molecule define the barbed end of the actin filament, while the pointed end is composed of Subdomains II and IV of the monomer (Dominguez, 2004; Kabsch et al., 1990; Kabsch and Vandekerckhove, 1992). Thus, in the cellular context, the (+) end usually points to the cell surface and the (-) end points inwards, away from the membrane (Dickinson and Purich, 2007). Multiple components of the actin monomers are involved in the actin-actin contacts that occur in the F-actin structure. The four subdomains of G-actin are thought to rearrange when it forms filaments. Each monomer in the protofilament has stronger interactions with its two closest neighbors, as well as weaker interactions with other monomers in the neighborhood. From the F-actin model, the amino acids of the actin monomers at these contact sites have been identified (Holmes et al., 1990; Lorenz et al., 1993). The hydrophobic cleft is thought to be involved in such inter-subunit contacts in F-actin. Existing evidence suggests that, in F-actin, the region defined as the D-loop (His40-Gly48) of an actin monomer binds in the hydrophobic cleft of a neighbouring monomer (Dominguez, 2004; Holmes et al., 1990). Labeling or cleavage of the D-loop affects actin polymerization (Burtnick, 1984; Khaitlina and Strzelecka-Golaszewska, 2002; Schwyter et al., 1990) Actin Turnover The actin cytoskeleton is very dynamic, undergoing rapid remodeling. Maintaining the actin network in a dynamic flux with regulated turnover of individual filaments! G

10 would allow the cells to remodel their cytoskeleton swiftly in response to internal and external cues. Remodeling of the actin filament system in cells results from strictly regulated polymerization and depolymerization of actin (Moseley and Goode, 2006). Reduced rates of filament turnover deplete the monomer pool and limit growth (Lappalainen and Drubin, 1997). The concentration of the ATP bound G-actin pool in the cytoplasm regulates the formation of F-actin structures and the availability of ATP bound G-Actin is governed by several steps. Actin turnover refers to the collective dynamic events of actin subunits treadmilling through filaments and dissociating from filament ends. These monomers undergo nucleotide exchange and are hence recycled into a pool of assembly competent ATP-bound G-actin available for new rounds of polymerization (Moseley and Goode, 2006) (See Illustration 2). G-actin- ATP Polymerization F-actin- ATP F-actin- ADP Depolymerization G-actin- ADP ADP/ATP exchange G-actin- ATP Pi ATP-Actin Monomer Severing/ ADF - eg. Cofilin ADP.Pi-Actin Monomer Nucleotide Exchange - eg. Profilin +! Actin filament ADP-Actin Monomer Monomer Sequestering/ CAPs Actin Nucleator Illustration 2. Actin turnover. Actin filaments are a mosaic enriched in ATP bound actin near the barbed (+) end, ADP-Pi actin in the middle and ADP-actin at the pointed (#) end. Cofilin binds to and severs older (ADP-bound) actin filaments leading to filament disassembly from the pointed end. The pool of ADP-Actin monomers undergo rapid nucleotide exchange to become ATP-Actin monomers. ATP-actin monomers get added to the barbed ends of filaments through the activity of actin nucleators. The ATP is hydrolyzed to actin bound to ADP and Pi and in a subsequent step the Pi is released from the subunit leaving behind actin bound to ADP. Many monomer sequestering proteins and CAPs regulate the entire process.! $K

11 Several proteins are important to ensure the rapid turnover of actin. As important as it is to assemble actin filaments, the rapid disassembly of filaments and recharging of G-actin bound ADP is equally crucial. Actin-depolymerizing factor (ADF)/cofilin is a well-conserved actin-modulating protein, which induces reorganization of the actin cytoskeleton by severing and depolymerizing F-actin (Lappalainen and Drubin, 1997). ADF/cofilin also binds to G-actin and inhibits nucleotide exchange, and hence, is supposed to regulate the nucleotide-bound state of the cellular G-actin pool cooperating with profilin (Shiozaki et al., 2009). Cyclase-associated proteins (CAPs) are also known to be involved in recycling G-actin monomers from ADF/cofilins for subsequent rounds of filament assembly (Bertling et al., 2004). Profilin another wellconserved G-actin-binding protein promotes nucleotide exchange (Haarer et al., 1990; Lu and Pollard, 2001). Thus, a number of actin monomer binding proteins co-operate to replenish the assembly competent pool of actin monomers by catalyzing the conversion of ADP monomers dissociating from ends of filaments to ATP-monomers. A defect in any of these proteins could lead to the sequestering of G-actin and could eventually affect actin polymerization/ depolymerization cycles, actin filament turnover and assembly of F-actin structures (Haarer et al., 1990; Lappalainen and Drubin, 1997). These different factors, their activities and mechanisms are described in the following section Actin Binding Proteins Actin polymerization is a classic example of self-assembly (Pollard, 2007), but cells tightly regulate all aspects of actin assembly by using a variety of proteins commonly termed Actin Binding Proteins (ABPs). Regulation is an essential aspect because the cytoplasm contains a high concentration of actin subunits that are available for! $$

12 polymerization and this pool of assembly-ready actin is controlled in many ways. A plethora of highly conserved actin-associated proteins and numerous upstream signaling molecules control the spatial and temporal assembly of actin structures (dos Remedios et al., 2003). These proteins direct the location, rate, and timing for actin assembly into different cytoskeletal networks. Typically, ABPs are multidomain proteins containing, in addition to their actin-binding domains, signaling domains and protein-protein interaction modules. The ABPs regulate actin nucleation, barbed end addition, pointed end dissociation, filament stability and nucleotide exchange on monomers. This in turn ensures the dynamic turnover of actin structures such that cells can rapidly alter their cytoskeletons in response to internal and external cues. Although ABPs are extremely diverse, both structurally and functionally, they mostly seem to share a common binding area on the actin surface in the cleft between actin subdomains 1 and 3 (Dominguez, 2004). This target-binding cleft of actin falls near the hinge for domain motions in actin. Thus, binding in this area is an effective way for ABPs to sense the conformation of actin, in particular conformational changes resulting from ATP hydrolysis by actin or from the G- to F-actin transition (Dominguez, 2000). Although the number of ABPs is extremely large and is constantly growing, the actin-binding domains of most ABPs can be grouped into structurally conserved folding families, including the WASP homology domain-2 (WH2) (Paunola et al., 2002), the actin-depolymerizing factor/cofilin (ADF/cofilin) domain (Lappalainen et al., 1998), the gelsolin-homology domain (McGough et al., 2003), the calponin-homology (CH) domain (Gimona et al., 2002), the formin homology 2 (FH2) domain (Higgs, 2005) and the myosin motor domain (Sellers, 2000), to mention just a few. The structures of complexes of actin with some members of these folding families are now known, and are starting to reveal common! $5

13 features in the way ABPs interact with actin. The main classes of ABP found in eukaryotic cells and their mechanisms of action are highlighted in this section with particular reference to the yeast actin cytoskeleton. Actin Nucleators Polymerization of actin is energetically unfavourable until three actin monomers associate to form a nucleus. This stage of filament formation is denoted the lag phase in vitro. In vivo, ABPs essentially function to remove the lag period and these are crucial to ensure the rapid nucleation of filaments. New filaments can form de novo, from the side of existing filaments, or by severing an existing filament. Different proteins function to promote each of these modes of nucleation. The first actin assembly factor to be identified was the Arp2/3 complex, which is a class of actin nucleators composed of seven polypeptides (ARPC 1 5, Arp2 and Arp3). The Arp2/3 complex works differently from the other class of actin nucleatorsthe formins. Arp2 and Arp3 are actin-related proteins having a similar tertiary structure to that of actin (Kelleher et al., 1995). The two molecules themselves are thought to mimic an actin dimer and elongate towards the barbed end (Higgs and Pollard, 2001). When the Arp2/3 complex binds G-actin the effect is to generate a stable trimer thereby serving as a nucleation site for the growth of a filament. The entire complex remains at the pointed end of the filament acting as a pointed-endcapping protein that encourages rapid growth of the filament from its barbed end (Goley and Welch, 2006). In addition, the Arp2/3 complex binds to the side of a preexisting actin filament, nucleating actin filaments from the side. This gives rise to a branched filament structure with an angle of 70 between the filaments (Amann and Pollard, 2001; Mullins et al., 1998). Such branched filaments are important to allow the dendritic branching that is found at the leading edge of motile cells (Pollard and! $9

14 Borisy, 2003). The functional interactions with other proteins, known as Nucleation Promoting Factors (NPFs), most importantly the WASP and SCAR/WAVE proteins, is likely to enhance the activity of the Arp2/3 complex in vivo (Goley and Welch, 2006). The G- actin-binding WASP homology 2 (WH2) domain of WASP is important for the actin-nucleating activity of Arp2/3 and its role is probably to feed actin monomers to the filament-barbed end. Verprolin/WIP binds both monomers and F-actin and has been shown to enhance Arp2/3-mediated actin polymerization when bound to the Arp2/3 activator cortactin (Paavilainen et al., 2004). The formins are a major class of actin filament nucleators that were identified rather recently. In contrast with the Arp2/3 complex, formins are single polypeptide, multidomain proteins (Higgs, 2005). All formins studied to date are dimeric, due to dimerization of their formin homology 2 (FH2) domain. The FH2 domain is responsible for driving actin nucleation of linear actin filaments. Studies indicate that the formin homology 2 (FH2) domain dimer flexes to accommodate the processive addition of actin monomers to the barbed end of a filament (Xu et al., 2004). After nucleation, the FH2 domain remains bound at the barbed end and moves along the filament as it elongates, promoting elongation by preventing the access of capping proteins. The association of profilin to the FH1 domain of formin enhances this formin-mediated elongation of actin filaments. The Rho family of GTPases is known to regulate formins through direct activation (Wallar and Alberts, 2003). Spire proteins are a third class of actin nucleators that were initially identified in Drosophila (Quinlan et al., 2005). Spire proteins are single polypeptides characterized by the presence of four Wasp homology 2 (WH2) motifs, which are responsible for actin nucleation (Quinlan et al., 2005). As with formins, spire-nucleated filaments are not branched.! $6

15 Monomer Binding Proteins The rapid growth of actin filaments that can be observed in motile cells and the reorganization of actin that occurs in response to both intracellular and extracellular cues in other cell types requires the availability of actin monomers to be tightly regulated. A typical total cytosolic concentration of actin in an animal cell is about 0.5 mm and with the Cc for G-actin being 0.1 µm, nearly all intra-cellular actin should exist as filaments with very little G-actin (Lodish, 2000). But from actual measurements about 40% of actin in animal cells are unpolymerized. The most likely explanation for this is that a group of highly conserved actin-monomer-binding proteins sequester actin in the cytosol to keep its G-actin concentration above the Cc yet holding it in a form that is unable to polymerize (Sun et al., 1995). Although there are a large number of monomer-binding proteins (>25 in mammalian cells alone), there are six major classes that are found in organisms from yeast to human and four of these classes are reported in plants (Winder and Ayscough, 2005). As a whole, the monomer-binding proteins are involved in binding ADP- actin as it is released from filament ends (e.g. twinfilin, ADF/cofilin), facilitating the nucleotide exchange of ADP for ATP (e.g. profilin and CAP) and delivering the monomer to barbed ends (or to Arp2/3) to facilitate new rounds of polymerization (e.g. twinfilin, Srv2/CAP, profilin, verprolin/wip and WASP). Monomer-sequestering proteins, the best studied of which is the thymosin family, act by clamping ATP- actin top to bottom. They effectively cap at both barbed and pointed ends, preventing incorporation into filaments (Hertzog et al., 2004; Irobi et al., 2004). Appropriate signals at the cell cortex can then trigger activation of profilin, leading to a rapid release of thymosin binding, resulting in a massive increase in the amount of actin available for polymerization.! $F

16 Filament Length Regulators Once nucleated, actin filaments are able to grow rapidly by addition of monomers at their barbed ends. However, filaments are regulated by several mechanisms. Capping proteins control the length of a filament. Barbed end capping proteins such as gelsolin and tensin, block addition of new monomers, leading to a decrease in the overall length of the filament. Pointed end capping proteins, by contrast, reduce loss of monomer from the pointed end thereby leading to rapid filament extension. In addition, gelsolin can sever actin filaments, thereby rapidly enhance actin filament recycling (Burtnick et al., 2004; Southwick, 2000). Depending on the environment in the cell this can have various outcomes, but it is usually a mechanism to disassemble F-actin- containing structures. Filament Severing & Depolymerizing Proteins The rapid turnover of actin filaments inside living cells is largely promoted by a group of small (15 22 kd) actin binding proteins collectively called the ADF/cofilin family (Abe et al., 1996). It includes cofilin, destrin, depactin, actophorin, and actindepolymerizing factor (ADF) of which cofilin is a ubiquitous and highly conserved protein. It binds to ADP-F-actin and promotes dissociation of ADP-actin from the pointed end of the actin filament. ADF/cofilin also associates with AIP-1 (actininteracting protein1). This interaction appears to increase the depolymerizing activity of cofilin. The complex is also thought to promote barbed end capping (Paavilainen et al., 2004). Filament Stabilizing Proteins Another important and highly conserved protein belonging to the family of ABPs is Tropomyosin. It binds along the filament length and, like other proteins that bind along the side of actin filaments, stabilizes the filament against spontaneous! $M

17 depolymerization (Broschat et al., 1989; Kojima et al., 1994). Furthermore, tropomyosin has a protective effect against gelsolin severing and ADF/cofilinmediated depolymerization. In concert with troponins, tropomyosin also plays an important role in regulating the interaction of myosin with the actin filament in striated muscles. Cortactin is another example that promotes and stabilizes Arp2/3-induced actin filament network formation (Weed et al., 2000). It is involved at the leading edge of migrating cells to drive protrusion. Cortactin is also thought to weakly activate the Arp2/3 complex (Weaver et al., 2001). Nebulin is another protein that associates along the length of filaments acting as rulers. It regulates thin filament architecture by a mechanism that includes stabilizing the filaments and preventing actin depolymerization (Pappas et al., 2010). Nebulin is an elongated protein with numerous low-affinity actin- binding sites and is thought to have a molecular ruler function in determining filament length. The length of the nebulin protein in different tissues and species corresponds with the length of the thin filaments in each half sarcomere. F-actin Bundling and Cross-Linking Proteins All higher-order F-actin structures are formed by two broad actin-binding activities; F-actin bundling and F-actin crosslinking. The organization of actin into higher-order structures is crucial to cells as these structures are responsible for the overall shape and order in a cell. For example, transient or dynamic structures such as the filopodia at the leading edge of migrating cells or the acrosomal processes of some invertebrate spermatozoa require actin to be organized into higher order networks. The microvilli of brush border epithelial cells and the architecture of striated muscles are other examples of higher order networks of actin structures.! $P

18 Actin bundling is the parallel or antiparallel alignment of F-actin into linear arrays. Either multimeric proteins that contain one actin-binding domain per subunit or proteins with two discrete actin-binding domains within their sequence are required to bundle actin. Actin bundles are usually further subdivided into loose or tight bundles, the topography of which is more or less dependent on the architecture of the bundling protein. The presence of two actin-binding domains in close proximity, as that in the protein fimbrin (Otto, 1994), leads to the formation of tight actin bundles. The more loosely ordered structures of actin stress fibres, however, are organized by the dimeric and antiparallel protein!-actinin, which has a single actin-binding site per subunit. These are separated by a helical spacer region, placing the two actin-binding domains of the dimer some distance apart, which results in a looser association of actin filaments. Both these proteins are known to have overlapping functions (Wu et al., 2001). Actin crosslinking proteins arrange actin filaments into orthogonal arrays, which are mediated by proteins or protein complexes that contain multiple actin-binding domains. In general, longer and more flexible spacer regions separate the two actinbinding domains, allowing a more perpendicular arrangement of actin filaments. The large flexible dimeric filamin or tetrameric spectrin complexes are examples of proteins of this type. Small monomeric proteins, such as transgelin, can also crosslink under certain conditions and organize actin filaments into dense meshworks. Interestingly, the yeast homologue of transgelin, known as Scp1p, induces tight actin bundles rather than meshworks (Otto, 1994; Winder et al., 2003). Cytoskeletal linkers and membrane anchors The utility of F-actin as a structural framework within cells necessitates its connection to other cellular elements. While some ABPs may indirectly alter actin dynamics and! $R

19 structure, their primary role is to anchor actin to membrane complexes or serve as a link between actin and other cytoskeletal elements. Some of these proteins, denoted as sidebinders and signalers, are a diverse group of ABPs that bind specifically to F-actin. In addition to their actin-binding domain, these proteins have domains that allow them to interact with other proteins. More specifically these proteins function within signaling networks to allow remodeling of the actin cytoskeleton at the appropriate time and place within the cell. For example, proteins like annexins directly bind to membranes and also interact with actin while several other proteins that contain poly-proline motifs (e.g. VASP and vinculin) recruit components of the actin polymerization machinery, such as profilin. Proteins such as dystrophin and utrophin or talin and vinculin, connect the actin cytoskeleton to the cell adhesion receptors dystroglycan or integrin, respectively; whereas several SH3-domain-containing proteins are involved in membrane trafficking (e.g. Abp1) and association with adhesion complexes (e.g. cortactin). Another important group of proteins link actin to other cytoskeletal elements such as microtubules and intermediate filaments. Plectin is one such protein that links actin to both microtubules and intermediate filaments. Clearly such proteins are of great importance to the cell in the integration of structure and signaling between the cytoskeletal elements and the maintenance of cell integrity (Winder and Ayscough, 2005). Motor Proteins An enormous number of ABPs use actin as a scaffold, physical support or track rather than affecting actin dynamics or regulating actin structure. Myosins are a large family of ATPases that hydrolyse ATP to generate the force to move along actin filaments driving a wide variety of cellular functions (Cope et al., 1996; DePina and Langford,! $G

20 1999). In other words, actin provides the tracks on which myosin can move, with ATP being the required fuel. There are over 17 different classes of myosins that use actin as a track to achieve various functions. Myosins move their specific cargo- be it vesicles, actin filaments or a host of other cellular components, mostly but not exclusively, from the pointed to the barbed end of the filament (Hodge and Cope, 2000) Post-translational modification of actin Actins undergo many types of post-translational modifications including arginylation, glutathionylation, and phosphorylation (Terman and Kashina, 2013).! Methylation of His-73 and removal of one or two residues from the N-terminus followed by acetylation occur commonly in most actins (Pollard and Cooper, 1986). There exists actin-specific processing enzymes due to the fact that actins are unusual in the type of amino acids that occur at its N-terminus and (Redman and Rubenstein, 1981; Rubenstein et al., 1981; Sheff and Rubenstein, 1992). The inhibition of such N- terminal processing in vitro is thought to alter actin polymerization (Hennessey et al., 1991). For example, arginylation increases actin polymerization and strengthens the actin filament network (Saha et al., 2010). However, yeast cells lack part of the N- terminal processing activity and completely blocking all N-terminal processing does not affect the stability of actins expressed in yeast (Cook et al., 1991). The current understanding of how these post-translational modifications exactly affect actin and its functions is still unclear Actin isoforms and isoform specific functions Actin is found in all eukaryotes and plays a fundamental role in many diverse and dynamic cellular processes. A mere incomplete list of actin functions include cell division, migration, junction formation, chromatin remodeling, transcriptional! 5K

21 regulation, vesicle trafficking and cell shape regulation (Chhabra and Higgs, 2007). An unresolved mystery is how one molecule accomplishes such a numerous list of tasks. The fact that actin in reality is composed of several different isoforms could thus provide a possible answer to this question with the idea being different isoforms perform distinct cellular functions (Perrin and Ervasti, 2010). The actin isoforms fall under three main groups: alpha, beta and gamma. Generally, alpha is present in muscles while beta and gamma actins coexist in most cell types as components of the cytoskeleton and as mediators of internal cell motility intracellular transport, cell shape maintenance and mitosis (Dugina et al., 2009; Perrin and Ervasti, 2010). The muscle actins are restricted to tissues with high tonic activity such as striated heart muscle, skeletal muscle or smooth muscle of blood vessels, gut wall and the urogenital system (Tondeleir et al., 2009). "-isoform preferably localizes in stress fibers, circular bundles and at cell-cell contacts as an unbranched filamentous array; $-actin displays a more variable distribution that is dependent on cellular activities. It is mainly organized as a branched meshwork with cortical and lamellar localization in moving cells, but also colocalizes with "-actin in lamellipodia or is recruited into stress fibres. In mammals, six actin isoforms can be distinguished: three!-actin isoforms (!-skeletal muscle,!-cardiac muscle and!-vascular), one "-isoform ("- cytoplasmic) and two $-isoforms ($-cytoplasmic and $-smooth muscle). Actin isoforms exhibit an unusually high degree of amino acid sequence similarity across species from algae to humans, indicating a strong conservation, despite actin being so ubiquitous (Meagher, 1991). With each isoform containing ~375 amino acids, only subtle sequence variations distinguish the isoactins. The amino acid sequence of!-actin differs from cytoplasmic actin isoforms in more than 20 residues! 5$

22 that are spread over the entire molecule. By contrast, differences between "- and $- actin are restricted to the N-terminus Asp2-Asp3-Asp4-Ile5 ("-actin) and Glu2- Glu3-Glu4-Ile5 ($-actin). The N-terminal sequences of!-skeletal and!-cardiac actin correspond to Asp3-Glu4-Asp5-Glu6 and Asp3-Asp4-Glu5-Glu6 (Vandekerckhove and Weber, 1978). These actin isoforms however, cannot completely replace each other in vivo and show marked differences in their tissue-specific and subcellular localization despite their remarkable sequence identity (Fyrberg et al., 1998; Kaech et al., 1997; Kumar et al., 1997; Mounier et al., 1997). Most species have several actin genes with plants having the largest numbers where a multi-gene family encodes the different isoforms (Baird and Meagher, 1987; Meagher, 1990). In certain cases such as Taenia solium (Schwob and Martin, 1992) and C. elegans (Krause et al., 1989) multiple actin genes encode the same amino acid sequence. Actins from different species appear to be more similar than those within a multi-gene family, again suggesting the possibility of isoform specific functions. For example, rice actin1 is more similar to Arabidopsis actin than any other rice actin (McElroy et al., 1990). Another example is the similarity between vertebrate cytoplasmic actins to arthropod actins rather than to vertebrate muscle actins (Hightower and Meagher, 1986). Within arthropods, Drosophila melanogaster and Bombyx mori muscle actins form a group distinct from their cytoplasmic actins (Mounier N et al, J. Mol. Evol. 34, 1992). Studies have proposed that conserved amino acids define the different groups of actin in vertebrates (Miwa et al., 1991), insect muscle (Mounier et al., 1992) and plants (Nairn et al., 1988). The amino acids at which mammalian skeletal muscle actin differs from " and $ actin occur as clusters in the 3D atomic structure (Kabsch and Vandekerckhove, 1992).! 55

23 A number of lower eukaryotes such as Aspergilus nidulans (Fidel et al., 1988) Saccharomyces cerevisiae (Lees-Miller et al., 1992) and the protozoan, Tetrahymena thermophila (Cupples and Pearlman, 1986) have only a single actin gene though some contain genes for actin-like proteins. For example, in addition to its single actin, Act1 (375 amino acids), S.pombe has an Arp3 protein (427 amino acids), which is 52.6% diverged from its actin (Lees-Miller et al., 1992). It is known from genetic experiments that Act1 cannot substitute for Arp3 and vice-versa demonstrating that, these two genes are essential for cell viability and have different functions (Lees- Miller et al., 1992). The idea that different isoforms of actin might perform distinct cellular functions might be explained by two possible mechanisms: i. A subset of actin-binding proteins could bind specifically to a single isoform and that specific interaction could be regulating its function. Consistent with this idea, several actin binding proteins have been shown to differentiate between muscle and cytoplasmic actin isoforms (De La Cruz, 2005; Larsson and Lindberg, 1988; Weber et al., 1992) ii. The subcellular localization of the different isoforms could be distinct resulting again in differential interactions with actin binding proteins. Such distinct localization patterns have also been observed in cell types. For example in neurons cultured from X. laevis, " cyto -actin protein is required for the normal turning behavior and the transcripts of the " cyto -actin accumulate on the side of a growth cone when exposed to an attractive cue (Leung et al., 2006; Yao et al., 2006). However, enrichment of $ cyto -actin on the side of the growth cone was not detected. Consistent with the targeting of " cyto -actin transcripts, several groups report that " cyto -actin is enriched at the leading edge! 59

24 of cultured fibroblasts and myoblasts as compared to stress fibres in the central region of the cell (Hill and Gunning, 1993; Hoock et al., 1991; Shestakova et al., 2001). In contrast, $ cyto -actin appears uniformly distributed in all actin-containing structures in fibroblasts (Otey et al., 1986) F-actin assemblies and in vivo functions of actin Actin was first isolated as a major muscle protein in the 1940s and by the 1970s, it slowly became apparent that cytoskeletal actin also played important roles unrelated to muscle function (Pollard and Cooper, 1986; Way and Weeds, 1990). Non-muscle actin is involved in processes such as cell and intracellular motility, cell division, and dynamic remodeling of the cytoskeleton. F-actin acts as an ATPase and actin filaments are also known to serve as molecular tracks for myosin motors to move along (Hennessey et al., 1993). Actin filaments mediate cell adhesion to the substratum by binding to focal contacts at the cell s lower surface. Migrating cells exert force on and pull themselves along these focal contacts. A diagnostic feature of the actin cytoskeleton of fibroblasts grown in culture is the so-called stress fiber system that traverses the cell near the plasma membrane facing the substratum. Stress fibers are loose bundles of actin-containing microfilaments probably stabilized and/or held together by bundling or crosslinking proteins. Stress fibers are typically nm wide and 5 10 µm long. Thus, within the cell individual microfilaments form a loose meshwork, whereas along the cell cortex they bundle into stress fibers (SF) that frequently appear to be associated with microtubules. Moreover, actin filaments might also play a role in compartmentalizing metabolic pathways within cells (Matsudaira, 1991). Last, but not least, the bacterial pathogen, Listeria monocytogene induces actin polymerization into polar rocket-tail structures within the host cell as a means to! 56

25 promote its rapid intracellular movement and its cell-to-cell transfer (Southwick and Purich, 1994). The cortical actin filaments in animal cells are organized into three general types of arrays (Bretscher, 1991). Parallel bundles, such as those found in microspikes and filopodia, where the filaments are oriented with the same polarity and are often closely spaced (10-20 nm apart). Contractile bundles, such as those found in stress fibers and in the contractile ring that divides cells in two during mitosis, where filaments are arranged with opposite polarities; these F-actin bundles are more loosely spaced (30-60 nm apart) and contain the motor protein myosin-ii. Gel-like networks of the cell cortex, where the filaments are arranged in a relatively loose, open array with many orthogonal interconnections (Bretscher, 1991) Understanding actin function and interactions through mutations The relationship between the function of a protein and its amino acid sequence can be studied by examining the effects of sequence variation. Molecular genetic techniques such as gene cloning and sequencing have accelerated the accumulation of new actin sequences in the databases. These natural actin isoforms can be used to assess the significance of amino acid sequence variation for functional differences between actins. Actin mutants are very useful in elucidating the roles of actin and actin binding proteins in all aspects of cell biology. Actin mutants have been recovered serendipitously from melanoma cells and transformed into tissue culture cells (Leavitt and Kakunaga, 1980; Taniguchi et al., 1986). Mutants obtained by selection for flightlessness or uncoordinated movement in Drosophila melanogaster and Caenorhabditis elegans, have led to the recovery of mutations in actin (Karlik et al., 1984; Kenyon, 1988; Mogami and Hotta, 1981). However, these are rather inefficient! 5F

26 ways to recover new actin mutants. The serendipitous recovery of actin mutants is rather difficult because severe actin mutations are often lethal. Many mutant actins also display antimorphic or dominant negative phenotypes in vivo (Sakai et al., 1990), i.e., the heterozygous organism exhibits the mutant phenotype (Sparrow et al., 1991b). Since filament formation involves several binding sites on the actin molecule, it is possible that a point mutation destroys only one of them with the remaining binding sites on this molecule still being functional. Therefore, the mutant actin molecule may nonproductively interact with wild-type molecules. The incorporation of perhaps only one mutant monomer may be sufficient to cap and prevent growth of an actin filament even in the presence of a large excess of wild-type monomers. This antimorphic behavior thus makes it difficult to interpret the in vivo effects of mutant actins. Although comparisons of isoform sequences and naturally occurring actin mutants are a valuable source of information about actin, unraveling the role of individual amino acids in actin function requires further investigation. Information on the importance of specific amino acids in actin can be obtained from artificially-induced mutations recovered from genetic studies (Hennessey et al., 1993). A number of mutant actins, obtained by either selection after random mutagenesis, genetic identification and isolation, or by mutagenesis in vitro have been produced (Hennessey et al., 1993; Sheterline et al., 1995). The in vitro mutagenesis of cloned actin genes permit any amino acid substitution to be made at any point in actin. They allow probing predictions of atomic models for the structure of the actin filament e.g., (Lorenz et al., 1993) and dissecting the anatomy of the actin molecule functionally both in vivo and in vitro. Despite the high sequence conservation and the well-known labiality of actin in vitro, it is possible with mutants to explore one part of the molecule without! 5M

27 destroying others. A major advantage of studying these mutations in vivo is that, the mutant actin can be analyzed under the normal environment where it assembles into cellular structures and interacts with a whole range of actin binding proteins. Amino acids directly involved in ATP and metal ion binding have been mutated (Solomon et al., 1988). Some other studies directly looked at the ability of mutants to bind ATP. In some cases, mutants that had substitutions in sites away from the ATP contact sites showed reduced ATP binding. This suggested that a change in the structural conformation in these mutated actins affected the ability to bind ATP (Drummond et al., 1992; Kabsch et al., 1990). The Glycine at residue 244 has been shown to be involved in actin-actin contacts and its mutation has been shown to affect polymerization (Leavitt et al., 1987; Taniguchi et al., 1988). Actins containing residue substitutions at other positions (P38A and C374S) have also been shown to have a reduced ability to polymerize (Aspenstrom et al., 1993; Taniguchi et al., 1988). Combining the atomic structure of actin and the extensive biochemical data of actin, made it possible to study the effect of specific actin mutations on its interactions with other proteins. These analyses led to the identification of the potential binding sites for many of the actin binding proteins known today (Kabsch and Vandekerckhove, 1992). The phenotypic effects of actin mutants may be diverse given the variety of actin functions and the different organisms used. However, the highly conserved nature of actin implies that mutants with effects in one organism are likely to have effects in another. Model organisms to study actin Actin is a rather challenging protein to study owing to its multifunctional nature in vivo. A variety of phenotypes can be seen in actin mutant cells due to the involvement! 5P