Update on Myosin Motors: Molecular Mechanisms and Physiological Functions^,

Abstract

Recent progress has revealed aspects of the molecular mechanisms that allow myosin motors to carry outtheir physiological functions.

The cytoskeleton is the main organizing principle that defines not only the distribution of cellular components within cells but also the overall shape and growth pattern of cells and entire multicellular organisms. This function is particularly evident in plant cells, where, for example, the organization of cortical microtubules determines the orientation of cellulose microfibrils in the growing cell wall and, hence, the direction of cell expansion and, ultimately, organ and plant shape (Szymanski and Cosgrove, 2009). Over the years, it has become clear that many functions of the cytoskeleton in eukaryotic cells depend heavily on motor proteins that use the energy stored in ATP to generate mechanical force along the length of cytoskeletal elements (Vale, 2003). This is also true for plant cells that use both microtubule-based kinesin motors and actin-based myosin motors to transport various cargoes along these cytoskeletal filaments as well as to modify the organization of these same filaments (Madison and Nebenführ, 2013; Geitmann and Nebenführ, 2015).

Myosin motors have emerged as powerful players that drive cytoplasmic streaming, actin organization, and cell expansion. These cellular functions allow myosins to perform supporting roles in plant growth, environmental responses, and defense against pathogens. Angiosperms typically encode around 15 myosin genes that can be grouped into two families, myosin VIII and myosin XI motors, which are found only in the Archaeplastidae lineage (Reddy and Day, 2001; Odronitz and Kollmar, 2007; Kollmar and Mühlhausen, 2017). Both motor families follow the typical myosin domain organization with an N-terminal motor domain, followed by a neck region that functions as a lever arm, a coiled-coil region for dimerization, and a globular C-terminal tail. Myosin XI is similar to myosin V in animals (Kinkema and Schiefelbein, 1994) due to the length of the neck region, which allows for 35-nm steps along actin filaments (Tominaga et al., 2003), and the DIL domain-containing globular tail, which functions in cargo binding (Li and Nebenführ, 2007). Myosin XI has been identified as the motor that propels organelles during cytoplasmic streaming that can reach speeds of over 10 µm s⁻¹ (Peremyslov et al., 2008, 2010; Prokhnevsky et al., 2008; Ueda et al., 2010; Madison et al., 2015). In angiosperms, the myosin XI motor family has evolved into up to eight subtypes (Myo11A–Myo11H; see Text Box 1) with a variable number of isoforms per group (Mühlhausen and Kollmar, 2013). Evidence for the functional specialization of these subtypes is still limited. Myosin VIII motors are smaller in size than myosin XI and localize primarily to the cell surface, where they are thought to function in endocytosis and plasmodesmatal trafficking (Reichelt et al., 1999; Golomb et al., 2008; Sattarzadeh et al., 2008; Haraguchi et al., 2014). In angiosperms, myosin VIII has evolved into at least two subtypes, but the functional relevance of this separation has not been addressed so far.

This Update provides a brief overview of recent progress in our understanding of both the molecular mechanisms and physiological functions of myosin motor proteins in plants.

NOT ALL MOTORS ARE CREATED EQUAL: DIFFERENCES IN ENZYMATIC PROPERTIES SET THE STAGE FOR SPECIALIZED FUNCTIONS

Myosin motors, like all cytoskeletal motors, couple the hydrolysis of ATP to a reversible conformational change in their motor domain, which is then translated into a larger movement by the stiff neck region at the C-terminal end of the motor domain (Preller and Manstein, 2013). This hydrolytic cycle is coordinated with the binding and release of an actin filament such that the motor protein takes one step toward the plus end of the filament with every ATP molecule used. The rate at which the hydrolytic cycle occurs combined with the length of the neck region defines the speed with which these motors can move along an actin filament. If two or more motors alternate in their binding to actin filaments, these motor dimers (or oligomers) can move processively along the filament and displace their cargo over long distances. Of course, this processive movement of myosin XI dimers (Li and Nebenführ, 2008) is only possible as long as each motor is bound to the filament for more than half of the time needed to complete a hydrolytic cycle (i.e. its duty ratio is greater than 0.5, so that the motor-cargo complex cannot diffuse away). In principle, the duty ratio can drop below 0.5 if several myosin dimers on an organelle surface can cooperate in moving the organelle (Krementsova et al., 2017), but it is not known whether this occurs in plant cells. Determination of the enzymatic properties of the motors encoded in plant genomes, therefore, should allow for a basic understanding of their possible functions within cells.

Probably the most impressive list of motor properties was established for myosin XI of Chara corallina. This motor can hydrolyze 400 molecules of ATP per second, which, combined with its long neck region (Ito et al., 2007), allows for translational speeds of 60 µm s⁻¹ (Higashi-Fujime et al., 1995). These extremely fast turnover rates are consistent with the high speed of cytoplasmic streaming, around 60 µm s⁻¹, found in this organism (Kamiya and Kuroda, 1956), suggesting that myosin XI is responsible for this stunning example of intracellular motility.

Several publications have described the enzymatic properties of different myosin XI motors, mostly from Arabidopsis (Arabidopsis thaliana) but also from tobacco (Nicotiana tabacum). Interestingly, these properties vary substantially within this motor family, raising the possibility that the different myosin XI isoforms have been selected to perform specific transport functions. For example, the turnover rate of ATP hydrolysis varies from 3.7 to 77 s⁻¹, allowing for a range of predicted speeds from 0.25 to 9 µm s⁻¹ (Yokota et al., 1999; Tominaga et al., 2003; Haraguchi et al., 2016). While the faster motor speeds are consistent with the maximal speeds of organelle movements described for various plant cells (Sieberer and Emons, 2000), the slower motors likely perform different functions, thus establishing a clear functional separation of the different myosin isoforms. Interestingly, the highly expressed Arabidopsis myosin XI genes Myo11B2 (MYA2), Myo11F (MYA1), and Myo11G (XI-I) that are disrupted in the frequently used 4ko quadruple knockout represent the full spectrum of myosin speeds. Myo11G is a slow motor (0.25 µm s⁻¹; Haraguchi et al., 2016), Myo11F in an intermediate motor (3.2 µm s⁻¹; Hachikubo et al., 2007), while Myo11B2 is fast (7.5 µm s⁻¹; Tominaga et al., 2013). The Myo11E (XIK) motor, which is disrupted by the fourth mutation in the 4ko line and which is often considered the most influential motor (Peremyslov et al., 2012), has not been characterized so far. It is not known whether the strong phenotype of the quadruple mutant stems from this combined loss of motors that function at different speeds and possibly move different cargo or whether it is simply a result of the low expression levels of the remaining myosins.

A crucial parameter that determines enzymatic turnover rates is the release of ADP from the nucleotide pocket, which allows, for example, the superfast myosin XI motors in C. corallina to reach such high hydrolysis rates (Higashi-Fujime et al., 1995; Ito et al., 2007). Since ADP-bound myosin binds actin more tightly than ATP-bound myosin, this low affinity for ADP comes at the price of a very brief tight binding between myosin and actin. As a result, the duty ratio of the motor is only 0.3 and dimers of myosin XI in C. corallina are not processive (Ito et al., 2007). Interestingly, a 175-kD myosin XI in tobacco is still processive despite rapid ADP release by slowing down the motor isomerization after initial binding of ATP, which maintains the motor in its tightly bound state longer (Diensthuber et al., 2015).

A critical question emerging from these results is how myosin isoforms can achieve such drastically different kinetic properties despite using a highly conserved motor domain. Targeted mutagenesis of C. corallina motors has revealed that the charge distribution between two surface loops within the actin-binding domain of the motor influences the turnover rate by modulating the rate of ADP release (Ito et al., 2009). These results were confirmed by direct experimental modification of a Dictyostelium discoideum myosin (Ito et al., 2009) and also are consistent with a comparison of structural homology models of different myosin motors (Diensthuber et al., 2015). It has been proposed that subtle differences in charge distributions in these two loops, as well as other places within the actin-binding region of myosin XI, also can explain the increased processivity of tobacco myosin XI (Diensthuber et al., 2015), but experimental evidence for this prediction is still missing. Better characterization of the structure-function relationship of myosin motor activities across the spectrum of motor activity will allow us not only to predict the motor properties of uncharacterized myosins but also to manipulate the functions of individual isoforms in a targeted way.

SLOW MYOSINS MAY FUNCTION MORE AS TENSORS THAN AS MOTORS

The realization that some plant myosins are relatively slow motors raises the question of which functions these motors perform. Specifically, slow motors are unlikely to contribute to fast organelle movements, as has been confirmed for Myo11G (Peremyslov et al., 2010). Interestingly, the slow motors Myo11G and Myo8C (ATM1) of Arabidopsis also have a high affinity for actin (Haraguchi et al., 2014, 2016). This property suggests that these motors may be more prone to holding onto filaments and generating (or responding to) tension in the actin network than to functioning as motors that move actively along filaments. This may explain the role of Myo11G in maintaining the elongated shape of plant nuclei (Tamura et al., 2013). The reduced mobility of nuclei in myo11g (kaku1) mutants (Tamura et al., 2013) may then just be a secondary effect of the reduced association of nuclei with actin filaments.

Myosin VIII seems to play a similar stabilizing role during cytokinesis in the moss Physcomitrella patens. In dividing protonemal cells, Myo8A localizes in a spot-like pattern to the cortical division site as well as to the plus ends of phragmoplast microtubules (Wu and Bezanilla, 2014). In a mutant that lacks all five myosin VIII genes, the phragmoplast fails to properly orient toward the cortical division site, leading to oblique cell walls (Wu and Bezanilla, 2014). It is thought that the cortical Myo8 motors generate tension on actin filaments that are formed by the formin For2a on the cell plate (van Gisbergen et al., 2012), leading to an actin array that defines the division plane. At the same time, the other population of Myo8 helps to guide microtubules along these actin filaments into the phragmoplast array (Wu and Bezanilla, 2014). A prediction derived from this model is that mutants with a loss of the cortical Myo8 localization should have defects in phragmoplast guidance even if localization to microtubules was not affected, but this has not been addressed experimentally.

Interestingly, it is not clear whether this model also applies to angiosperm cells. While it has been shown that myosin VIII also localizes to the cortical division site prior to mitosis (Wu and Bezanilla, 2014) and to the cell plate region during cytokinesis (Van Damme et al., 2004), no clear defect in plant growth was detected in quadruple knockouts in Arabidopsis that had lost all their myosin VIII genes (Talts et al., 2016). This lack of an apparent mutant phenotype contrasts with the clear morphological defects resulting from global myosin VIII loss in P. patens (Wu et al., 2011; Wu and Bezanilla, 2014). This discrepancy may suggest that the preprophase band of microtubules found in angiosperm cells (Müller, 2012) functions redundantly with the Myo8-based steering mechanism. Combinations of myosin VIII mutants with mutants affected in the preprophase band-associated mechanism should help to identify possible overlapping functions. Given the localization of full-length Myo8C to actin filaments in interphase cells (Haraguchi et al., 2014), it also should be interesting to investigate possible changes in the actin cytoskeleton in the quadruple Myo8 mutant.

INTERACTIONS BETWEEN MYOSINS AND CARGO PROTEINS SUGGEST LIMITED ISOFORM SPECIFICITY

A promising approach to discerning the molecular functions of individual motor isoforms is to identify the cellular components that these myosins act upon by isolating interacting proteins that may function as adaptors for larger cargo complexes or organelles. In the case of myosin XI, the cargo-binding domain was identified as the C-terminal globular tail domain based on homology to the related myosin V motors in yeast and animals (Li and Nebenführ, 2007). Systematic yeast two-hybrid screens for proteins that interact with this domain resulted in the identification of several protein families (Fig. 1). The best-characterized among these groups is the MyoB family that is defined by the presence of a domain of unknown function, DUF593 (Peremyslov et al., 2013). MyoB proteins form large families in both dicots (Arabidopsis) and monocots (Oryza sativa), and subfamilies are predicted to carry various additional domains in addition to DUF593, such as transmembrane domains, coiled-coil regions, or metal-binding regions (Peremyslov et al., 2013). The localization of fluorescent protein-tagged MyoB proteins revealed their presence on numerous small, punctate structures that moved rapidly through cells in a pattern that was very similar to Myo11E-labeled spots (Peremyslov et al., 2013; Kurth et al., 2017). A tobacco homolog of the MyoB family, RISAP, was found associated with the trans-Golgi network in pollen tubes (Stephan et al., 2014), but none of the tested Arabidopsis MyoB proteins colocalized with a variety of established organelle markers in Arabidopsis (Peremyslov et al., 2013; Kurth et al., 2017). It remains to be seen whether these MyoB compartments define a novel set of endomembrane vesicles that were previously unknown or whether they represent subdomains of known organelles. In this context, it is interesting that Myo11E can bind MyoB proteins and is required for their normal distribution and motility (Peremyslov et al., 2013). In addition, Myo11E was identified previously in density gradient fractions together with membranes of the endoplasmic reticulum (ER; Peremyslov et al., 2012) and also plays a major role in ER motility (Ueda et al., 2010). It is tempting to speculate that MyoB proteins define small subdomains on the ER surface that function as motility hubs by recruiting clusters of myosin motors (Fig. 1).

Figure 1.

Interactions of myosins, myosin-binding proteins, and organelles. Various myosin XI isoforms were found to interact with different classes of binding proteins. The identity of the cargo has not been identified for several of the binding proteins (question marks in black circles). Myosin VIII is found in the cell cortex, where it may function in creating tension in the actin network. For details, see text. ER, Endoplasmic reticulum; P-body, processing boy; TGN, trans-Golgi network.

In addition to the MyoB family, additional myosin XI-interacting proteins have been identified. They include two groups of proteins, MadA and MadB, that lack a transmembrane domain and, therefore, are assumed to function as adaptors between motors and other proteins on organelle surfaces (Kurth et al., 2017). Furthermore, an interaction between myosin XI and DECAPPING PROTEIN1 (DCP1; Steffens et al., 2014), a central component of processing bodies that are centers of RNA processing in cells (Bailey-Serres et al., 2009), was detected. Finally, two proteins on the nuclear envelope, WIT1 and WIT2 (for WPP DOMAIN-INTERACTING TAIL-ANCHORED PROTEIN1 and -2), were identified as interacting with Myo11G, the myosin responsible for the elongated nuclear shape and movements in vegetative cells (Tamura et al., 2013).

This large number of myosin XI-interacting proteins might suggest a series of specific interactions that allow the motors to operate on subsets of cargoes. Curiously, this does not seem to be the case. For example, a recent broad analysis of pairwise interactions by yeast two-hybrid assay revealed remarkably little specificity in binding among five Myo11 and 10 MyoB isoforms (Kurth et al., 2017). A similar analysis for MadA proteins showed some more specificity, but at least one family member, MadA1, interacted nonselectively with all tested myosins (Kurth et al., 2017). A similar lack of isoform specificity was found for the interaction between myosin XI and DCP1 (Steffens et al., 2014). In this case, all tested motors spanning a range of myosin XI subtypes were found to associate with DCP1 under a variety of experimental conditions, including bimolecular fluorescence complementation in plant cells (Steffens et al., 2014). The only case of specific binding between a motor and a cargo protein described to date is the interaction between Myo11G and WIT1/2 (Tamura et al., 2013; Kurth et al., 2017), suggesting that this early branching myosin XI has evolved a unique function in nuclear shape and movement. This interpretation also is consistent with Myo11G’s very slow ATP hydrolysis cycle (Haraguchi et al., 2016). This unique function of Myo11G, however, seems to be redundant with another myosin in pollen tubes, since wit1 wit2 double mutants show defects in nuclear movement whereas mutants in myo11g (kaku1-4) do not (Zhou and Meier, 2014), demonstrating that other myosin(s) also may be able to bind to WIT1/2.

Taken together, the growing body of research on myosin-cargo interactions paints a complex picture with limited specificity of motors for the different cargos in plant cells (Fig. 1). It will be interesting to test whether this apparent promiscuity also can be detected in plant cells when several different proteins compete for myosin XI binding or whether differences in binding affinities lead to a sorting of different myosin XI isoforms to different cargos. In the former case, it will be necessary to investigate how the presence of several isoforms with inherently different motor activities (see above) affect cargo motility. In the latter case, it will be important to dissect the mechanisms that lead to motor segregation such as differences in protein structure or posttranslational modifications.

MYOSIN MOTORS FUNCTION MOSTLY REDUNDANTLY IN A VARIETY OF GROWTH PROCESSES

An important question regarding cytoskeletal motors is the effect of their molecular and cellular functions on the physiology of the plant. In the case of myosin XI, it became clear from mutant analysis in Arabidopsis that the activity of at least some isoforms is necessary for cell expansion. This was most evident in root hairs and pollen tubes that depend heavily on the delivery of secretory vesicles to the growing tip of these cells (Ojangu et al., 2007; Peremyslov et al., 2008; Park and Nebenführ, 2013; Madison et al., 2015). Combinations of several myosin XI mutants resulted in broader effects and progressively smaller plants that were made up of smaller cells in a wide range of tissues (Prokhnevsky et al., 2008; Peremyslov et al., 2010; Ojangu et al., 2012), suggesting the largely redundant function of these myosins in cell expansion. An elegant experiment in transgenic Arabidopsis plants further established that the speed of myosin-driven movements plays a pivotal role in plant growth. In this case, a myo11b2 mutant was complemented with either the wild-type Myo11B2 gene or with one where the motor domain had been replaced with that of the faster myosin XI from C. corallina (see above) or the slower myosin Vb from humans (Tominaga et al., 2013). Interestingly, the growth rates of the resulting plants correlated roughly with the speed of the motors and with displacement rates of cytoplasmic streaming, demonstrating a clear relationship between intracellular motility and plant growth (Tominaga et al., 2013).

How the activity of myosin motors is translated into faster growth is not known at this time. Two candidates for this mechanism are the active delivery of secretory vesicles to the cell surface to directly support growth (Crowell et al., 2010) and the general mixing of cytoplasm, which is assumed to stimulate metabolism (Shimmen and Yokota, 1994). In addition, it is possible that the effect of myosin on growth is more indirect, for example by affecting actin filament dynamics (Park and Nebenführ, 2013; Cai et al., 2014) and organization (Peremyslov et al., 2010; Ueda et al., 2010; Madison et al., 2015). While there seems to be some experimental support for the first model (Cramer and Mitchison, 1995; Gutierrez et al., 2009; Sampathkumar et al., 2013), the second remains largely unexplored. In addition, the insight that organelle speeds seem to correlate with growth rates raises other questions. For example, does the acceleration of one motor (Myo11B2) lead to an increase in movement rates of specific organelles needed for growth? Why are both faster and slower versions of Myo11B2 epistatic over the remaining endogenous isoforms? Could an acceleration of relatively slow motors, such as Myo11F and Myo11G, also boost growth rates? Answers to these questions not only will allow us to better define the cellular functions of myosin motors but also will provide a better understanding of the mechanisms by which myosins contribute to growth.

Given that myosin motor activity is required for growth, it may not be astonishing that a growth response like gravitropism is affected in higher order myosin XI mutants (Talts et al., 2016). Specifically, the loss of Myo11B2, Myo11E, and Myo11F resulted in plants that had reduced stem growth rates, and these stems were stiffer and bent more slowly during gravistimulation (Talts et al., 2016). Interestingly, however, the sedimentation of amyloplasts was slower in the triple mutants (Talts et al., 2016), suggesting that their reduced gravitropism cannot be ascribed solely to slower growth. Previous research has established that amyloplasts are remarkably dynamic in endodermis cells (Saito et al., 2005), and some of their movements are reminiscent of myosin-driven organelle movement. It is conceivable that myosins mediate at least some of the dynamics of amyloplasts and, thereby, contribute to gravity perception.

Curiously, myosin motors also were found to contribute to the straightening reaction that occurs at the end of gravitropic or phototropic bending. In this case, a myo11e myo11h (xik xif) mutant of Arabidopsis was found to show initially normal gravitropic bending but then overshot beyond a vertical orientation (Okamoto et al., 2015). Experiments with plants on klinostats demonstrated that the mutant defect only concerns the straightening reaction, suggesting that these two myosins are crucial in reestablishing symmetrical growth once the stem reaches an orientation close to vertical. Interestingly, Myo11H is expressed almost exclusively in a thin layer of fiber cell surrounding the xylem cells (Okamoto et al., 2015), indicating that this cell layer is particularly important in establishing straight growth. These fiber cells contain long actin filament bundles and also show rapid movement of plastids. While the plastid movements are much slower in the myosin mutant (Okamoto et al., 2015), we do not know whether the organization of actin filaments also is affected. This latter point is of particular importance, since we know that myosins can affect actin organization (Peremyslov et al., 2010; Ueda et al., 2010; Madison et al., 2015) and since fiz1 mutants that are defective in ACTIN8 also show similar defects in straightening behavior (Okamoto et al., 2015). More fundamentally, it is not clear at this time whether these two myosins function in sensing the growth curvature, possibly via the actin cytoskeleton, or whether the mutant phenotype results from a defect in the response pathway after detection of the growth curvature.

PLANT-PATHOGEN INTERACTIONS REVEAL NEW FUNCTIONS OF MYOSIN MOTORS

Viruses are opportunistic pathogens that routinely subvert endogenous mechanisms and pathways of their host cells for their own ends. This is also the case for the acto-myosin system that is used by viruses to accelerate their systemic spread across tissues, as demonstrated by a series of experiments in Nicotiana benthamiana. For example, silencing of Myo11B in this species resulted in a slower spread of GFP-labeled Tobacco mosaic virus across the leaf epidermis (Harries et al., 2009). The supporting role of myosins in viral spread also was confirmed with experiments using dominant negative constructs where overexpression of the motor-less myosin tails blocks normal motor function. While the precise mechanism of this effect is not known (Vick and Nebenführ, 2012), this approach was used successfully to reduce the mobility of several organelles (Avisar et al., 2008, 2009, 2012; Sparkes et al., 2008, 2009; Griffing et al., 2014). Similarly, the cell-to-cell spread of Tobacco mosaic virus and Grapevine fanleaf virus in N. benthamiana leaf epidermis could be reduced by overexpressing Myo11B and Myo11E tail constructs (Amari et al., 2011, 2014), again confirming a role of these motors in the transport of viruses. Interestingly, a similar myosin-dependent movement was detected for the VirE2 protein of Agrobacterium tumefaciens that is injected into plant cells together with the T-DNA (Yang et al., 2017). At this point, it is not clear whether the viruses or VirE2 interact directly with myosin motors or whether they are moved passively together with endogenous organelles, since these experimental interventions also reduced organelle motility. Irrespective of the precise mechanism involved, however, both types of pathogens appear to exploit the intracellular transport machinery in plant cells to their own ends. This implies that myosin mutants should be less susceptible to viral infection, but this has not been tested so far.

Interestingly, the opposite effect was found for a fungal pathogen, where myosin mutants were more susceptible to the pathogen. Specifically, myosin XI knockouts in Arabidopsis were more readily colonized by several biotrophic ascomycete pathogens and even showed symptoms with Colleotrichum destructivum, which is normally not able to penetrate wild-type cells (Yang et al., 2014). These phenotypes were most pronounced in triple and quadruple myosin XI mutants but were already evident in single mutants that lacked the highly expressed Myo11B2 or Myo11E gene (Yang et al., 2014). The quadruple myosin mutant further showed defects in callose and lignin deposition as well as PEN1/SYP121 levels (Yang et al., 2014), which had been shown previously to be involved in successful defense against biotrophic pathogens (Hardham et al., 2007). Fungal pathogens are known to trigger reorganization of the actin cytoskeleton as well as the accumulation of several organelles in the area of the penetration site (Takemoto et al., 2003). Since myosins can increase actin dynamics (Park and Nebenführ, 2013; Cai et al., 2014) and affect actin organization (Peremyslov et al., 2010; Ueda et al., 2010; Madison et al., 2015), it is reasonable to assume that this actin rearrangement upon fungal attack is compromised in the myosin mutants. Taken together, these observations suggest that similar mechanisms may be involved in fungal pathogen resistance to those in cellular expansion (see above).

Curiously, a recent article presented data suggesting that some biotrophic pathogens have subverted part of the acto-myosin system to aid in the penetration of plant cells. In this case, the oomycete Hyaloperonospora arabidopsidis showed reduced infectivity in mutants that lack MyoB1 (Schroeder et al., 2016), indicating that this myosin interactor plays an important role in the infection process. Interestingly, the pathogen locally induced the expression of MyoB1, presumably to support its own growth on the plant host. This novel function of MyoB1 seems to contradict the role of myosins in defending against ascomycete pathogens, but it is not known whether this function is related to MyoB1’s role in binding myosin motors to cargo membranes. In this context, it is important to note that the resistant phenotype of myob1 mutants manifests itself already as a single mutant (Schroeder et al., 2016), whereas growth defects are detectable only in the case of multiple MyoB knockouts (Peremyslov et al., 2013), suggesting that the involvement in pathogen responses is unique for MyoB1. It will be interesting to determine the cellular defect(s) in myob1 that allows H. arabidopsidis to colonize Arabidopsis more readily and whether any myosin mutants also show a similar phenotype.

CONCLUDING REMARKS

Recent years have seen steady progress in the characterization of myosins in plants. The availability of null mutants in Arabidopsis has enabled the discovery of the physiological functions of myosins that allow plants to grow or respond to their environment more efficiently. At the same time, characterization of the enzymatic properties and identification of tail-binding proteins has allowed us to peer deeper into the mechanisms that are employed by myosins to carry out their cellular functions. While these advances help us to paint a more complete picture of plant myosin functions, it is also clear that there are still many empty spots on our canvas where we lack clear data (see Outstanding Questions). A combination of biochemical, advanced cellular, and genetic approaches will be necessary to tackle the complexities of the acto-myosin system in plant cells.