63 Cytoskeletal control of plant cell shape: getting the ®ne points Laurie G Smith The shapes of plant cells, which are de®ned by their surrounding walls, are often important for cell function. The cytoskeleton plays key roles in determining plant cell shape, mainly by in¯uencing the patterns in which wall materials are deposited in expanding cells. Studies employing cytoskeleton-disrupting drugs, together with studies of mutants with cytoskeletal defects, have demonstrated that both microtubules and actin ®laments are critical for all modes of cell expansion, although their precise roles remain poorly understood. In recent years, however, signi®cant progress has been made in understanding the contributions of a variety of proteins that in¯uence cell shape by regulating the organization and polymerization of cytoskeletal ®laments in expanding cells. Addresses Section of Cell and Developmental Biology, Division of Biology, University of California San Diego, 9500 Gilman Drive, La Jolla, California 92093-0116, USA e-mail: lsmith@biomail.ucsd.edu Current Opinion in Plant Biology 2003, 6:63±73 This review comes from a themed issue on Growth and development Edited by Thomas Laux and John Bowman 1369-5266/03/$ ± see front matter ß 2003 Elsevier Science Ltd. All rights reserved. DOI 10.1016/S1369-5266(02)00012-2 Abbreviations ADF actin-depolymerizing factor AN ANGUSTIFOLIA Arp2,3 Actin-related protein 2,3 AtCAP1 Arabidopsis thaliana ADENYLYL CYCLASE-ASSOCIATED PROTEIN1 AtKTN1 Arabidopsis thaliana KATANIN1 BARS brefeldin A adenosine diphosphate-ribosylated substrate BOT1 BOTERO1 Brk Brick CA constitutively active CtBP carboxy-terminal binding protein F-actin ®lamentous actin FRA2 FRAGILE FIBER2 GFP green ¯uorescent protein KCBP KINESIN-LIKE CALMODULIN BINDING PROTEIN MAPK mitogen-activated protein kinase MOR1 MICROTUBULE ORGANIZATION1 ROP Rho of plants SIMK stress-induced MAPK SPK1 SPIKE1 TON2 TONNEAU2 ZWI ZWICHEL Introduction Plant cells exhibit a wide variety of shapes, which are often functionally signi®cant. For example, the highly www.current-opinion.com elongated shapes of root hairs, pollen tubes, trichomes and vascular elements are vitally important to their functions. Even subtle features of cell shape can have a signi®cant impact on function. For example, the lobed shapes of mesophyll cells are thought to enhance their capacity for gas exchange during photosynthesis, and the conical shapes of petal epidermal cells affect their optical properties so as to enhance coloration. The shapes of plant cells are de®ned by their walls, and are acquired during development according to the patterns in which walls expand during organ and cell growth. Cell expansion depends on the ability of the wall to yield in a controlled manner under the force of turgor pressure exerted by the cell within. This is achieved through constant breakage and re-formation of bonds between wall components combined with controlled deposition of new wall materials. Wall composition and structure are critical for cell shape determination, but the cytoskeleton also plays key roles in cell shaping, mainly because of its in¯uence on the pattern in which wall materials are deposited. This review highlights recent advances in our progress toward understanding how the cytoskeleton contributes to plant cell shape determination, and how the polymerization and organization of cytoskeletal ®laments are regulated in expanding cells. Diffuse growth Most cells expand diffusely, meaning that wall extension and the incorporation of new wall material are distributed across the cell surface. Diffuse growth is generally anisotropic in that it is oriented preferentially along one axis. This property is thought to be conferred mainly by the arrangement of cellulose micro®brils, which serve as the principle structural component of the walls of expanding cells. Cross-linked cellulose micro®brils constrain cell expansion, and their alignment along one axis favors growth in the perpendicular axis. Studies spanning four decades have demonstrated that the orientation of diffuse growth depends critically on microtubules. Like cellulose micro®brils, microtubules are generally aligned perpendicular to the major axis of cell expansion in diffusely growing cells (Figure 1a). The treatment of diffusely expanding cells with microtubuledisrupting drugs causes them to expand more isotropically (reviewed in [1]). Moreover, recent genetic studies have con®rmed a critical role for microtubules in orienting diffuse cell expansion. Constitutive expression of an antisense a-tubulin gene in transgenic Arabidopsis reduced both the mRNA levels of several a-tubulin isoforms and the overall level of a-tubulin protein. Shortly after germination, this resulted in root growth effects that closely resembled those produced by treatment with oryzalin, a Current Opinion in Plant Biology 2003, 6:63±73 64 Growth and development Figure 1 (a) (b) Cortical microtubule Cytoplasmic F-actin bundle Cortical F-actin filament Current Opinion in Plant Biology Schematic summary of the organization of microtubules and F-actin in expanding cells. (a) In a diffusely growing cell, microtubules are essentially restricted to the cell cortex and are aligned perpendicularly to the major axis of cell expansion. In this schematic, the microtubules are aligned transversely and cell expansion is oriented longitudinally. Cortical F-actin is co-aligned with microtubules, but cytoplasmic F-actin bundles are oriented longitudinally. This drawing is based on data for rapidly expanding cells in the elongation zone of maize roots [23,24]. (b) In a tip-growing cell (in this case, a root tip), both F-actin bundles and microtubules are longitudinally oriented. A meshwork of fine actin filaments is found in the sub-apical region. This drawing is based on data that are summarized in [28,29]. microtubule-disrupting drug [2]. An interesting insight into the effects of microtubules on the polarity of diffuse growth has recently come from analysis of two lefty mutations of Arabidopsis, which cause left-handed helical twisting of cell growth in roots and other organs [3]. Two different lefty mutant a-tubulin proteins have identical missense mutations. These proteins are incorporated into microtubules but destabilize them, causing the microtubules to adopt a helical arrangement that predicts the twisted growth pattern. Recently, genetic studies have led to the identi®cation of a few of the undoubtedly large number of proteins that play critical roles in determining the orientation of diffuse cell expansion by regulating microtubule dynamics or arrangement in diffusely expanding cells. The Arabidopsis TONNEAU2 (TON2) gene is required for normal spatial organization of cortical microtubules in expanding and dividing cells, and encodes a protein predicted to function as a regulatory subunit for a type-2A protein phosphatase [4]. Thus, understanding the function of TON2 in microtubule organization will probably depend on de®ning the substrates of the associated phosphatase. Arabidopsis thaliana KATANIN1 (AtKTN1) is a katanin-like protein that is encoded by the mutationally de®ned Current Opinion in Plant Biology 2003, 6:63±73 FRAGILE FIBER2 (FRA2), BOTERO1 (BOT1) and ECTOPIC ROOT HAIR3 (ERH3) genes [5±7]. Reduced growth anisotropy in the cells of fra2 and bot1 mutants is associated with abnormal microtubule organization [5,6]. As katanins have been shown to function as microtubulesevering proteins [8], work on AtKTN1 implicates microtubule severing as an important mechanism for achieving the proper organization of microtubules in expanding cells. The Arabidopsis MICROTUBULE ORGANIZATION1 (MOR1) gene encodes a protein that is homologous to structural microtubule-associated proteins (MAPs) of the highly conserved MAP215 class [9]. Shifting temperaturesensitive mor1-1 and mor1-2 mutants to restrictive temperatures causes the rapid fragmentation and disorganization of cortical microtubules in expanding cells of several tissues, which is associated with loss of growth anisotropy. Other mutant alleles of MOR1 reveal that this gene also plays an essential role in cytokinesis [10]. MOR1 protein binds to microtubules in vitro and co-localizes with microtubules at all stages of the cell cycle in vivo [10], con®rming its role in stabilizing/organizing microtubule arrays via direct association with microtubules. A question that remains to be answered fully is that of how microtubules in¯uence the orientation of diffuse cell growth. In a wide variety of plant species and cell types, cellulose micro®brils are deposited into the wall in a pattern mirroring that of cortical microtubules, and the pharmacological disruption of microtubules is often associated with changes in the pattern of cellulose deposition. These observations have led to the hypothesis that cortical microtubules determine the cellulose deposition pattern by guiding the movement of cellulose synthase complexes through the plasma membrane (e.g. [11]). Although often stated as an established fact, proof for this longstanding hypothesis has been elusive. The literature is sprinkled with observations that are not readily accommodated by this hypothesis; for example, cellulose micro®brils can sometimes align locally in the absence of microtubules (reviewed in [12]). Alternative models have been proposed to explain the alignment of cellulose micro®brils, but these models do not account for the in¯uence that microtubules appear to have on this process (e.g. [13]). A recent, comprehensive review of this topic culminated in a model for cellulose micro®bril alignment through `templated incorporation' that can account for a wide variety of relevant observations [12]. Illustrated in Figure 2, this model proposes that nascent cellulose micro®brils can be aligned by their binding interactions with existing micro®brils, and can also be oriented through binding interactions with plasma membrane-associated proteins that are linked directly or indirectly to microtubules on the opposite site of the membrane. To account for changes in cellulose deposition pattern that are often observed to accompany the re-orientation of microtubules, the model proposes that the microtubule-based system generally takes precedence when guidance provided by www.current-opinion.com Cell shape Smith 65 Figure 2 (a) Nascent microfibril PM (b) PM MT Membrane proteins Scaffold Wall matrix Current Opinion in Plant Biology Model for the orientation of cellulose microfibrils in the cell wall by `templated incorporation' [12]. (a) The orientation of a nascent cellulose microfibril is specified by a scaffold that binds to an extant microfibril. (b) The orientation of the nascent cellulose microfibril is specified by a scaffold that binds to plasma membrane (PM) proteins, which are oriented with respect to a cortical microtubule (MT). For simplicity, plasma membrane proteins are shown binding the microtubule directly but there may be intermediaries. Although wall-dependent and microtubule-dependent scaffolds are drawn separately here, the model proposes that both scaffolds are usually present simultaneously and cooperate to orient nascent microfibrils. Redrawn from [12] with permission. existing micro®brils con¯icts with that provided by microtubules. This could be explained if the binding interactions between nascent micro®brils and microtubulelinked guidance molecules were tighter than those between nascent micro®brils and existing micro®brils. Finally, it must be recognized that microtubules may have more than one role in diffuse cell expansion. Indeed, growth alterations have occasionally been associated with the loss or altered organization of microtubules apparently without corresponding alterations in the pattern of cellulose deposition (e.g. [14±16]). Given that wall components besides cellulose and callose are introduced via secretion, another possible role for cortical microtubules in diffuse growth is the local guidance of secretory vesicles to appropriate sites in the plasma membrane. Evidence supporting this notion comes from analysis of Arabidopsis fra2 mutants, which lack the putative microtubule-severing katanin, AtKTN1. The walls of these mutants are de®cient in both cellulose and hemicellulose, a secreted wall component [5]. Although the role of microtubules in regulating diffuse cell expansion has received more attention than those of www.current-opinion.com other cytoskeletal components, it has become increasingly clear that actin ®laments also play an important role. Unlike microtubules, however, ®lamentous actin (Factin) seems to act mainly by promoting cell expansion per se rather than by controlling the pattern of expansion. This conclusion is based in part on the effects of actindisrupting drugs, which slow diffuse cell expansion in a variety of plant organs without altering the polarity of growth [17,18]. Moreover, altering the levels of proteins that regulate actin polymerization also causes changes in growth rates. Consistent with its known biochemical function, overexpression of actin-depolymerizing factor (ADF) in transgenic Arabidopsis caused a partial loss of Factin and reduced growth, whereas reduction of ADF levels through expression of an antisense version of the gene promoted the formation of excess F-actin and excess growth [19]. Overexpression of AtCAP1 (Arabidopsis thaliana ADENYLYL CYCLASE-ASSOCIATED PROTEIN1), which is related to animal and yeast CAPs that inhibit actin polymerization in vitro and in vivo, reduced cell expansion in transgenic tobacco plants [20]. Furthermore, overexpression of AtCAP1 essentially halted the growth of cultured tobacco Bright Yellow-2 Current Opinion in Plant Biology 2003, 6:63±73 66 Growth and development (BY-2) cells while causing a dramatic loss of F-actin in these cells [20]. Alterations in levels of pro®lin, an actinmonomer-binding protein that can either promote or inhibit actin polymerization depending on the experimental conditions, have had variable effects on cell growth. Nevertheless, these studies also broadly support the conclusion that F-actin has an impact on the rate of diffuse cell expansion [21,22]. Little is known about how the actin cytoskeleton promotes diffuse cell expansion. Unlike microtubules, which are essentially restricted to the cell cortex except during cell division, F-actin bundles permeate the cytoplasm (e.g. [23]; see Figure 1a). F-actin is also present in the cortex of expanding cells, where it tends to be co-aligned with microtubules (e.g. [24]; see Figure 1a). Actin-disrupting drugs can cause alterations in the arrangement of cortical microtubules (e.g. [24,25]), suggesting a role for cortical F-actin in organizing cortical microtubules. Given the different impacts of disrupting F-actin and microtubules, however, the effects of F-actin on microtubule organization are unlikely to completely explain the Factin's role in diffuse cell expansion. By analogy to a wellestablished role for F-actin in tip-growing cells (discussed later), a likely role for F-actin in diffuse growth is to transport secretory vesicles to the cell surface [19,26]. Tip growth Tip growth is a mode of polarized cell expansion in which wall extension and the incorporation of new wall material are focused at a single site on the cell surface. Thus, a tipgrowing cell elongates unidirectionally. Pollen tubes and root hairs are the only well-characterized tip-growing cell types in angiosperms. However, there may be other cell types that also employ this mode of elongation, at least some of the time, as suggested by the observation that cells in xylogenic suspension culture appear to expand via tip growth [27]. Actin-disrupting drugs either slow or stop tip growth (reviewed in [28]), not unlike their effects on diffusely growing cells. Actin ®laments are arranged in longitudinal bundles that run along the length of pollen tubes and root hairs (see Figure 1b). The myosin-driven movement of secretory vesicles along these actin cables transports them to the vicinity of the growth site, where their fusion adds new cell wall material and membrane [29]. However, Factin has long been suspected to play one or more additional roles in tip growth. This notion has recently been strengthened by the observation that tip growth is considerably more sensitive to a variety of actin-disrupting treatments than is the cytoplasmic streaming that drives long-range vesicle transport to the tip region [30]. Consideration of other possible roles for F-actin in tip growth has focused mainly on the population of ®laments near the growth site, where the majority of recent studies reveal a sub-apical network of ®ne actin ®laments (see Current Opinion in Plant Biology 2003, 6:63±73 Figure 1b). This network sometimes appears as a collar around the base of the growth site [31±33,34]. However, time-lapse analysis of living tobacco pollen tubes has demonstrated that this F-actin population is highly dynamic, and individual ®lament bundles transiently penetrate into the extreme tip [35]. A wide variety of models have been proposed to explain what F-actin might do to promote tip growth besides driving long-range vesicle transport via cytoplasmic streaming (reviewed [28,29]). Among these, the most widely favored ideas seem to be that actin ®laments near the tip are somehow involved in the local regulation of vesicle docking or fusion, or in facilitating endocytosis, which is needed to remove excess membrane. A variety of actin-binding proteins with de®ned biochemical functions are present in tip-growing cells and contribute to the regulation of actin dynamics (pro®lin, ADF) or actin bundling (villin; reviewed in [29]). A growing body of work demonstrates that Rac/Rho/ Cdc42-related small GTPases, which form a distinct family in plants (called ROPs [Rho of plants], with 11 members in Arabidopsis), also play critical roles in tip growth, including the regulation of actin dynamics. Arabidopsis ROP1 and the closely related ROP5 are enriched at pollen-tube tips and are required for the properly polarized growth of pollen tubes [36±38]. Functional studies of ROP1 have demonstrated that it plays roles both in the formation of the tip-high Ca2 gradient that is characteristic of tip-growing cells [37] and in the regulation of actin dynamics at the tip [35]. More recently, three other closely related members of the Arabidopsis ROP family, ROP2, ROP4 and ROP6, have been implicated in the polarization of root hair growth. Antibody localization showed that ROP proteins are enriched at root hair tips [39]. In particular, ROP2 is expressed in elongating root hairs and green ¯uorescent protein (GFP)::ROP2 is enriched in root tips [34]. The expression of constitutively active (CA) forms of ROP2, ROP4 and ROP6 caused the depolarization of root hair growth; in the case of CA-ROP4 and CA-ROP6, this was associated with delocalization of the calcium gradient [34,39]. The expression of dominant negative ROP2 caused the loss of the ®ne F-actin meshwork from the tip region of root hairs and greatly reduced root hair elongation [34]. Thus, ROP2, ROP4, and ROP6 are likely to promote properly polarized root hair growth by playing roles that are analogous to those of ROP1 and ROP5 in pollen tubes. Another recently identi®ed regulator of root hair growth with an actin-related function is the mitogen-activated protein kinase (MAPK), stress-induced MAPK (SIMK) [40]. This protein is enriched in root hair tips. The expression of a hyperactivated form of SIMK either accelerated or prolonged root hair growth, and treatment with the MAPK inhibitor UO 126 caused abnormal root hair growth. Studies using drugs have indicated that www.current-opinion.com Cell shape Smith 67 F-actin is necessary both for tip localization and for the activation of SIMK. Although the substrates of SIMK have not yet been identi®ed, a possible role for SIMK in polarized vesicle traf®cking at the tip was suggested by observations of aberrant vesicle behavior in living UO 126-treated hairs. Thus, SIMK could provide a link between F-actin and the regulation of exocytosis and/ or endocytosis at the tip. Roles for microtubules in tip growth have received relatively little attention, mainly because treatment of tipgrowing cells with microtubule-inhibiting drugs often shows little or no effect (reviewed in [28]). Recent work has shown, however, that treatment of growing Arabidopsis root hairs with microtubule-interacting drugs causes them to display a wavy growth pattern and to undergo branching [41]. Root hair branching was also observed in transgenic Arabidopsis plants that expressed an antisense a-tubulin gene [2]. These results suggest that microtubules play a role in stabilizing the growth site in Arabidopsis root hairs, but how they do this remains unclear. In tip-growing cells of angiosperms, microtubules are generally found in longitudinal, helical, or net-axial arrays. These arrays extend further into the tip in root hairs than in pollen tubes (reviewed in [28,29]; see Figure 1b). Given the proposed role of microtubules in diffusely growing cells, an interesting possibility is that microtubules may function to direct cellulose deposition at the growth site of root hairs. In contrast to pollen tubes, in which there is very little cellulose at the tip [29], a randomly oriented network of cellulose is found at the growth site of root hairs. However, Bibikova et al. [41] propose that the role of microtubules is to somehow restrict the location of the tip-high Ca2 gradient to a single site. The initiation of tip growth involves the selection of a growth site and the initial bulging of the cell wall at that site. It is distinct from tip growth proper, involving different mechanisms and different genes (reviewed in [42]). The initial bulging of root hairs is predicted by local changes in the pH of the cell wall [43], by the local accumulation of the cell wall-loosening enzymes expansin [44] and xyloglucan endotransglycosylase [45], and by the local enrichment of ROPs [34,39]. Surprisingly, the localization of most of these early indicators of bulge formation is insensitive to both actin- and microtubuledisrupting drugs, suggesting a cytoskeleton-independent mechanism for bulge site selection [39,44,45]. Nevertheless, actin ®laments become reoriented with respect to the bulge site during a very early stage of its formation [44]. Moreover, recent analysis of the phenotype resulting from a mutation in the Arabidopsis ACTIN2 gene clearly indicates a role for actin in selecting and focusing the bulge at the appropriate site on the surface of root hairforming cells [46]. A possible role for ROP2 in regulating the proper organization of F-actin at the bulge site is www.current-opinion.com indicated by the formation of multiple bulges on individual root hair-forming cells when wildtype ROP2 is overexpressed [34]. Earlier work had suggested an analogous role for ROP1 in establishing the site of tip growth in pollen grains [37]. Generation of complex cell shapes: trichomes and pavement cells Many plant cells have relatively simple shapes that are acquired through either diffuse or tip growth (e.g. rectangular or tubular shapes, respectively), whereas others have more complex shapes involving growth patterns that are less well characterized. Among these are epidermal trichomes that have branched shapes, such as those in Arabidopsis, and epidermal pavement cells that have marginal lobes, which are found almost universally among angiosperms (see Figure 3). The involvement of the cytoskeleton in the multidirectional cell expansion patterns that are involved in generating these complex cell shapes has received increasing attention over the past few years, and some interesting advances have been made recently. Super®cially, the outgrowth of unbranched trichomes resembles that of root hairs. However, studies employing microtubule-depolymerizing drugs, combined with examination of the arrangement of microtubules and wall polymers, have indicated that cotton ®bers (which are unbranched trichomes that emerge from the seed surface) grow via a highly polarized form of diffuse growth rather than by tip growth [47,48]. In line with these early ®ndings, more recent work has shown that the initial, polarized outgrowth of Arabidopsis trichomes and trichome branches is sensitive to microtubule-interacting drugs, which cause a more isotropic growth pattern [49,50]. Moreover, mutations in genes that encode putative microtubule/tubulin-interacting proteins interfere with the formation of trichome branches. These mutations include those affecting the AtKTN1 gene (which encodes a putative microtubule-severing katanin) [5], and weak mutations in the KIESEL and PORCINO genes (which encode tubulin-folding co-factors) [51,52]. As expected for a microtubule-dependent and polarized diffuse growth process, the emergence of trichomes and trichome branches involves the organization of microtubules into arrays with a net alignment transverse to the axis of elongation [53]. Although actin-disrupting drugs do not affect the initial, polarized outgrowth of trichomes and trichome branches, they severely inhibit the subsequent, rapid elongation of trichome branches [49,50]. During trichome branch elongation, actin ®laments are arranged in longitudinal bundles similar to those seen in tip-growing cells [49,50]. However, analysis of the growth pattern in elongating trichome branches reveals wall extension along the length of the branch, so branch elongation is not a tip growth process (M HuÈlskamp, personal communication). Thus, F-actin probably plays Current Opinion in Plant Biology 2003, 6:63±73 68 Growth and development similar roles in elongating trichome branches to those that it plays in other diffusely growing cell types undergoing rapid elongation. Figure 3 (a) More than a dozen genes are required for proper trichome morphogenesis in Arabidopsis (reviewed in [54]). Phenotypic analyses of mutants that are affected in some of these genes, combined with molecular analyses of the corresponding gene products, are beginning to shed light on the cytoskeleton-dependent mechanisms that govern trichome shape. The Arabidopsis ZWICHEL (ZWI) gene, which is required for the normal elongation of trichome stalks and the formation of the full complement of trichome branches, encodes a kinesin [55]. This kinesin has also been named KCBP (KINESIN-LIKE CALMODULIN-BINDING PROTEIN) and shown to have Ca2/ calmodulin-regulated, minus end-directed microtubule motor activity [56]. ZWI/KCBP functions in cell division, pollen tube growth, and pollen germination, as well as in trichome morphogenesis [57±59]. Although it remains possible that ZWI/KCBP functions in trichomes to transport an as yet unidenti®ed cargo to speci®c cellular sites, current information suggests that its function is probably to promote the formation of microtubule arrays that are critical for branch formation and stalk elongation. This conclusion is supported, albeit indirectly, by data implicating a role for KCBP in the formation or maintenance of microtubule arrays during cell division [59], and by the observation that treatment of developing zwi trichomes with taxol (a microtubule-stabilizing drug) can stimulate the formation of a normal number of trichome branches [53]. Localization of ZWI in expanding trichomes, and detailed comparison of microtubule organization in wildtype and zwi trichomes, might help to further elucidate the role of ZWI in trichome morphogenesis. (b) Recently, an intriguing connection has been made between ZWI and another gene that promotes trichome branch formation, ANGUSTIFOLIA (AN ). In addition to its function in trichomes, AN is also required for properly oriented growth of other leaf cell types [60]. Subtle defects in microtubule organization have been observed in the expanding leaf cells of an mutants, both trichomes and other cell types [61,62], indicating that AN could act directly at the level of microtubule organization to promote normally polarized cell expansion. The cloning of AN by two research groups showed that it encodes a protein that is related to CtBP (carboxy-terminal binding protein)/BARS (brefeldin A adenosine diphosphateribosylated substrate) proteins. These proteins have see- (c) Current Opinion in Plant Biology Scanning electron micrographs of complex cell shapes. (a) A typical, three-branched Arabidopsis trichome. (b) Arabidopsis epidermal pavement cells with large, irregular, lobes that interdigitate to form a jigsaw-puzzle-like arrangement. Stomata are interspersed among the Current Opinion in Plant Biology 2003, 6:63±73 pavement cells. (c) Maize epidermal pavement cells with small, regularly arranged, finger-like lobes that interdigitate to form a zipper-like arrangement. A cell file with three stomata is shown at the top of the figure. www.current-opinion.com Cell shape Smith 69 mingly divergent functions as transcriptional co-repressors and as proteins that are directly involved in the regulation of Golgi dynamics in animal cells [61,62]. AN::GFP fusion proteins are found in the nucleus and cytoplasm of onion epidermal cells [61] and predominantly in the nucleus of Arabidopsis cells [62], but the localization of AN::GFP in expanding trichomes has not been reported. Thus, the connection between AN and microtubules is not obvious from the sequence of this protein or from its localization. Nevertheless, there is some evidence that AN interacts directly with ZWI, and thus also with microtubules. Plants that are doubly heterozygous for certain alleles of an and zwi have abnormal trichome morphogenesis, and ZWI and AN have been shown to interact in the yeast two-hybrid system [61]. Further work will be required to determine how AN and ZWI may work together to organize microtubules for the proper formation of trichome branches. The morphogenesis of lobed epidermal pavement cells, like that of trichomes, involves multi-directional cell expansion that depends on both microtubules and Factin. Numerous studies have examined the organization of microtubules in lobe-forming cells in both the epidermis and mesophyll of various species. The emergence of lobes is consistently preceded by the reorganization of cortical microtubules into a series of bands that are associated with the formation of local, cellulosic wall thickenings (e.g. [63±66]). Lobes subsequently emerge between these microtubule bands/wall thickenings (see Figure 4a). These observations suggest that the thinner regions of the wall are more extensible than the thicker regions, and therefore that the thinner regions bulge out to form lobes as the cell expands under the force of turgor pressure. Further evidence of an essential role for microtubules in the formation of pavement cell lobes is provided by the failure to form lobes of the epidermal cells of the leaves of Arabidopsis fra2 mutants, which lack the putative microtubule-severing katanin AtKTN1 [5]. Until recently, the role of F-actin in lobe formation has received relatively little attention. One study showed that cortical F-actin was organized into bands that coincided with microtubule bands in expanding wheat mesophyll cells; drug treatments suggested a role for F-actin in the organization of microtubule bands [25]. However, recent studies have suggested a different role for F-actin in the formation of lobes in epidermal pavement cells. In both maize and Arabidopsis, local enrichments of cortical F-actin are found at the sites of lobe emergence, and these enrichments persist at lobe tips as they elongate ([67,68,69]; see Figure 4a). Thus, the organization of F-actin at lobe tips is reminiscent of that seen in tip-growing cells, except that the F-actin enrichment appears to extend further into Figure 4 (a) Wild type Microtubule F-actin (b) Brick mutant Current Opinion in Plant Biology A schematic summary of cytoskeletal organization in expanding epidermal pavement cells of a maize leaf. Microtubules are shown in black and actin in red. (a) In a wildtype cell, lobe formation is associated with the organization of microtubules into bands that are focused at lobe sinuses. In addition, localized enrichments of cortical F-actin are found at the sites of lobe emergence and at the tips of elongating lobes. (b) Pavement cells of brk1, brk2 and brk3 mutants expand to the same extent as wildtype pavement cells but without forming lobes. During cell expansion, the cortical microtubule bands that are associated with lobe formation in the wildtype are present (although somewhat less distinct than those in the wildtype), but localized enrichments of F-actin in the cell cortex are not observed. The drawing summarizes data from [67,68]. www.current-opinion.com Current Opinion in Plant Biology 2003, 6:63±73 70 Growth and development the extreme apex of emerging lobes than of root tips or pollen tubes. These observations raise the possibility that lobe formation involves a tip growth-like process that occurs at multiple sites along the cell margin, in addition to a microtubule-dependent process that helps to localize growth to speci®c sites. Whatever the role of actin may be, a variety of genetic studies have recently converged to support the conclusion that local F-actin polymerization is indeed critical for lobe formation and have shed new light on the molecular mechanisms involved. Mutations in the Arabidopsis SPIKE1 (SPK1) gene have a variety of effects on cell and organ growth, including loss of lobes from epidermal pavement cells and lack of trichome branching [70]. The spatial segregation of microtubule- and actin-enriched areas of the cell cortex found in expanding wildtype pavement cells is not found in the corresponding spk1 mutant cells, which instead have a uniform distribution of both classes of ®laments. SPK1 encodes a protein whose carboxy-terminal domain is related to the signature domain of CDM-family (CED5, DOCK180, MYOBLAST CITY family) proteins, which are implicated in reorganization of the cytoskeleton in animal cells in response to diverse extracellular cues [70]. Thus, the SPK1 protein may mediate the reorganization of the cytoskeleton in response to extracellular signals that govern the pattern of lobe formation. The CDM domain itself interacts with Rac GTPase, suggesting that SPK1 might interact with a ROP GTPase. Although various aspects of the spk1 mutant phenotype suggest that SPK1 acts primarily at the level of microtubule organization, Rac and ROP GTPases are better known for their involvement in regulating the actin cytoskeleton. An important question for future work is thus whether SPK1 directly regulates the organization of microtubules, F-actin, or both. If SPK1 does interact with a ROP GTPase, then a good candidate is Arabidopsis ROP2, discussed earlier in connection with the tip growth of root hairs. Interestingly, ROP2 also plays a crucial role in epidermal lobe formation [69]. Similar to its distribution in root hairs, GFP::ROP2 is enriched at the tips of emerging lobes. The expression of constitutively active ROP2 caused de-localization of cortical F-actin and a relatively uniform pattern of expansion, resulting in the formation of larger pavement cells that lacked lobes. The expression of dominant negative ROP2 inhibited the formation of localized enrichments of cortical F-actin and reduced the outgrowth of lobes [69]. Thus, these results suggest that ROP2 regulates lobe formation, at least in part, by activating the assembly of cortical F-actin in discrete regions of the cell cortex, a role similar to that played by ROP2 in root hairs. The possibility of a direct interaction between SPK1 and ROP2 will be an interesting one to explore in the future. In maize, three Brick genes (Brk1, Brk2 and Brk3) are required for the formation of lobes in epidermal paveCurrent Opinion in Plant Biology 2003, 6:63±73 ment cells [67,68]. The microtubule bands that are associated with lobe formation in wildtype expanding pavement cells are also present in the corresponding cells of all three brk mutants. However, local enrichments of cortical F-actin that are found in wildtype cells at the sites of lobe initiation and at the tips of emerging lobes fail to form in brk cells ([67,68]; see Figure 4b). These results support the conclusion that local enrichments of cortical F-actin play a crucial role in lobe formation in epidermal pavement cells, a role that extends beyond simply promoting the formation of microtubule bands. In addition, the results point to a role for the Brk gene products in stimulating the formation of the F-actin enrichments. Mosaic analysis and double-mutant analysis of brk mutations indicated that all three Brk genes act in a common pathway in which each gene has a distinct function [68]. The Brk1 gene encodes a novel 8-kDa protein that is highly conserved in both plants and animals [67]. Recent biochemical studies have implicated the mammalian homolog of BRK1 in the regulation of actin polymerization via its activation of the Arp2,3 (Actin-related protein 2,3) complex. Putative components of the Arp2,3 complex are encoded by predicted genes in various plant genomes. When activated by proteins that respond to localized intracellular or extracellular cues, this complex nucleates the polymerization of new actin ®laments at speci®c sites in the cell [71]. The putative mammalian ortholog of BRK1, HSPC300, is found in a multi-protein complex with the Arp2,3 activator WAVE, which is regulated by Rac [72]. Although the intact WAVE/ HSPC300-containing complex is inactive, WAVE and HSPC300 dissociate from the rest of the complex in the presence of GTP::Rac to form a two-protein subcomplex that activates Arp2,3-dependent actin polymerization. Thus, BRK1 most likely functions by associating with an activator of the Arp2,3 complex in plant cells. Interestingly, however, no proteins that are obviously related to WAVE can be identi®ed in the Arabidopsis genome. The hypothetical BRK1-associated activator cannot, therefore, be predicted on the basis of sequence but may be identi®able as a BRK1-interacting protein or may correspond to the product of the Brk2 or Brk3 gene. Another interesting possibility for future investigation is that the Arabidopsis BRK1 is part of a complex whose activity in stimulating local actin polymerization is regulated by ROP2 (or perhaps by another ROP). Conclusions Studies on the cytoskeletal regulation of plant cell morphogenesis have mainly focused on the role of microtubules in diffuse growth and F-actin in tip growth. It has become increasingly clear, however, that both diffuse and tip growth, along with the multidirectional growth patterns that are involved in the morphogenesis of trichomes and epidermal pavement cells, involve a collaboration between F-actin and microtubules. In recent years, a wide www.current-opinion.com Cell shape Smith 71 variety of proteins have been identi®ed that contribute to cell shape determination by regulating the organization and polymerization of cytoskeletal ®laments. Approaches that use a combination of genetic, cell biological, and biochemical tools are making signi®cant progress in understanding the functions of these proteins and their interactions with the cytoskeleton. However, the precise roles of cytoskeletal ®laments themselves in promoting appropriate patterns of cell expansion are still poorly understood. Genetic approaches have not yet helped much in this area, but may provide progress if mutants can be identi®ed that have cell morphogenesis defects that are not due to abnormal cytoskeletal organization or to loss of particular wall components. Subsequent analysis of the corresponding gene products might reveal functions in mediating interactions of the cytoskeleton with the cellulose-deposition machinery, the secretory system, Ca2 channels, and so on. Likewise, biochemical approaches that are aimed at identifying cytoskeletoninteracting proteins that are associated with the plasma membrane or the vesicles of expanding cells might be productive. In the years ahead, the availability of fully sequenced genomes for Arabidopsis and rice together with an ever-growing set of tools for reverse genetics, the analysis of protein±protein interactions, and advanced imaging methods will facilitate progress in all of the approaches used to improve our understanding of the cytoskeletal regulation of plant cell shape. Acknowledgements Thanks to Mary Frank, Heather Cartwright and many other colleagues for stimulating discussions over the years on the subject material of this review, to Martin HuÈlskamp for contributing information before publication, and to Tobias Baskin for Figure 2. Work on plant cell shape in the author's laboratory is supported by National Science Foundation grants IBN-9817084 and IBN-0212724. References and recommended reading Papers of particular interest, published within the annual period of review, have been highlighted as: of special interest of outstanding interest 1. Wymer C, Lloyd C: Dynamic microtubules: implications for cell wall patterns. Trends Plant Sci 1996, 7:222-228. 2. Bao Y, Kost B, Chua N-H: Reduced expression of a-tubulin genes in Arabidopsis thaliana speci®cally affects root hair development and root gravitropism. Plant J 2001, 28:145-157. 3. Thitamadee S, Tuchihara K, Hashimoto T: Microtubule basis for left-handed helical growth in Arabidopsis. Nature 2002, 417:193-196. Mis-sense mutations in Arabidopsis a-tubulin destabilize microtubule polymers, causing them to adopt a helical con®guration in expanding root cells and promoting left-handed helical twisting of root cell growth. The authors propose that the destabilization of microtubules might provide a general mechanism for helical growth in plants. 4. Camilleri C, Azimzadeh J, Pastuglia M, Bellini C, Grandjean O, Bouchez D: The Arabidopsis TONNEAU2 gene encodes a putative novel protein phosphatase 2A regulatory subunit essential for the control of the cortical cytoskeleton. Plant Cell 2002, 14:833-845. The Arabidopsis TON2 gene is found to encode a protein that is related to a type-B00 regulatory subunit of protein phosphatase 2A. This subunit can interact in the yeast two-hybrid system with an Arabidopsis type-A subunit of protein phosphatase 2A. www.current-opinion.com 5. Burk DH, Liu B, Zhong R, Morrison WH, Ye Z-H: A katanin-like protein regulates normal cell wall biosynthesis and cell elongation. Plant Cell 2001, 13:807-827. 6. Bichet A, Desnos T, Turner S, Grandjean O, HoÈfte H: BOTERO1 is required for normal orientation of cortical microtubules and anisotropic cell expansion in Arabidopsis. Plant J 2001, 25:137-148. 7. Webb M, Jouannic S, Foreman J, Linstead P, Dolan L: Cell speci®cation in the Arabidopsis root epidermis requires the activity of root hair 3, a katatin-p60 protein. Development 2002, 129:123-131. 8. McNally F, Vale RD: Identi®cation of katanin, an ATPase that severs and disassembles stable microtubules. Cell 1993, 75:419-429. 9. Whittington AT, Vugrek O, Wei KJ, Nasenbein NG, Sugimoto K, Rashbrooke MC, Wasteneys GO: MOR1 is essential for organizing cortical microtubules in plants. Nature 2001, 411:610-613. 10. Twell D, Park SK, Hawkins TJ, Schubert D, Schmidt R, Smertenko A, Hussey PJ: MOR1/GEM1 has an essential role in the plant-speci®c cytokinetic phragmoplast. Nat Cell Biol 2002, 4:711-714. 11. Giddings TH Jr, Staehelin LA: Microtubule-mediated control of micro®bril deposition: a re-examination of the hypothesis. In The Cytoskeletal Basis of Plant Growth and Form. Edited by Lloyd CW. London: Academic Press; 1991:85-99. 12. Baskin TI: On the alignment of cellulose micro®brils by cortical microtubules: a review and a model. Protoplasma 2001, 215:150-171. 13. Emons AMC, Mulder BM: How the deposition of cellulose micro®brils builds cell wall architecture. Trends Plant Sci 2000, 5:35-40. 14. Emons AMC, Wolters-Arts AMC, Traas JA, Derksen J: The effect of colchicine on microtubules and micro®brils in root hairs. Acta Bot Neerl 1990, 39:19-27. 15. Okuda K, Mizuta S: Modi®cation in cell shape unrelated to cellulose micro®bril orientation in growing thallus cells of Chaetomorpha moniligera. Plant Cell Physiol 1987, 28:461-473. 16. Sugimoto K: Cortical Microtubules, Cellulose Micro®brils and Growth Anisotropy in the Roots of Arabidopsis thaliana. Ph.D. thesis, Australian National University, Canberra; 2000. 17. Thimann KV, Reese K, Nachmias VT: Actin and the elongation of plant cells. Protoplasma 1992, 171:153-166. 18. Baluska F, Jasik J, Edelmann HG, Salajova T, Volkmann D: Latrunculin B-induced plant dwar®sm: plant cell elongation is F-actin-dependent. Dev Biol 2001, 231:113-124. 19. Dong C-H, Xia G-X, Hong T, Ramachandran S, Kost B, Chua N-H: ADF proteins are involved in the control of ¯owering and regulate F-actin organization, cell expansion, and organ growth in Arabidopsis. Plant Cell 2001, 13:1333-1346. 20. Barrero RA, Umeda M, Yamamura S, Uchimiya H: Arabidopsis CAP regulates the actin cytoskeleton necessary for plant cell elongation and division. Plant Cell 2002, 14:149-163. The authors demonstrate that overexpression of an Arabidopsis CAP (a regulatory subunit of adenylyl cyclase that is implicated in the regulation of actin dynamics in animals and yeast) causes loss of F-actin and slows or stops growth as well as cell division. 21. Ramachandran S, Christensen H, Ishimaru Y, Dong C-H, Wen C-M, Cleary AL, Chua N-H: Pro®lin plays a role in cell elongation, cell shape maintenance, and ¯owering in Arabidopsis. Plant Physiol 2000, 124:1637-1647. 22. McKinney EC, Kandasamy MK, Meagher RB: Small changes in the regulation of one Arabidopsis pro®lin isovariant, PRF1, alter seedling development. Plant Cell 2001, 13:1179-1191. 23. Baluska F, Vitha S, Barlow PW, Volkmann D: Rearrangements of F-actin arrays in growing cells of intact maize root apex tissues: a major developmental switch occurs in the postmitotic transition region. Eur J Cell Biol 1997, 72:113-121. Current Opinion in Plant Biology 2003, 6:63±73 72 Growth and development 24. Blanca¯or EB: Cortical actin ®laments potentially interact with cortical microtubules in regulating polarity of cell expansion in primary roots of maize ( Zea mays L.). J Plant Growth Regul 2000, 19:406-414. 25. Wernicke W, Jung G: Role of cytoskeleton in cell shaping of developing mesophyll of wheat (Triticum aestivum L.). Eur J Cell Biol 1992, 57:88-94. The mitogen-activated protein kinase SIMK is found to have a function in promoting the tip growth of root hairs. F-actin is required for both the proper localization and the function of this kinase. 41. Bibikova TN, Blanca¯or E, Gilroy S: Microtubules regulate tip growth and orientation in root hairs of Arabidopsis thaliana. Plant J 1999, 17:657-665. 26. Baskin TI, Bivens NJ: Stimulation of radial expansion in Arabidopsis roots by inhibitors of actomyosin and vesicle secretion but not by various inhibitors of metabolism. Planta 1995, 197:514-521. 42. Carol RJ, Dolan L: Building a hair: tip growth in Arabidopsis thaliana root hairs. Philos Trans R Soc Lond B Biol Sci 2002, 357:815-821. A valuable summary of recent studies on root hair tip growth that includes information about the roles of a wide variety of required genes. 27. Roberts AW, Uhnak KS: Tip growth in xylogenic suspension cultures of Zinnia elegans L.: implications for the relationship between cell shape and secondary-cell-wall pattern in tracheary elements. Protoplasma 1998, 204:103-113. 43. Bibikova TN, Blanca¯or EB, Gilroy S: Localized changes in apoplastic and cytoplasmic pH are associated with root hair development in Arabidopsis thaliana. Development 1998, 125:2925-2934. 28. Geitmann A, Emons AMC: The cytoskeleton in plant and fungal cell tip growth. J Microsc 2000, 198:218-245. 44. Baluska F, Salaj J, Mathur J, Braun M, Jasper F, Samaj J, Chua N-H, Barlow PW, Volkmann D: Root hair formation: F-actin-dependent tip growth is initiated by local assembly of pro®lin-supported F-actin meshworks accumulated within expansin-enriched bulges. Dev Biol 2000, 227:618-632. 29. Hepler PK, Vidali L, Cheung AY: Polarized cell growth in higher plants. Annu Rev Cell Dev Biol 2001, 17:159-187. A recent, comprehensive review of tip growth in pollen tubes and root hairs that integrates information about the roles of the cytoskeleton with other aspects of tip-growth regulation. 30. Vidali L, McKenna ST, Hepler PK: Actin polymerization is essential for pollen tube growth. Mol Biol Cell 2001, 12:2534-2545. A variety of F-actin-disrupting treatments inhibit pollen-tube tip growth at concentrations much lower than those required to inhibit cytoplasmic streaming. Thus, this study suggests that F-actin has other functions in tip growth besides long-range transport of secretory vesicles via cytoplasmic streaming. 31. Miller DD, Lancelle SA, Hepler PK: Actin micro®laments do not form a dense meshwork in Lilium longi¯orum pollen tube tips. Protoplasma 1996, 195:123-132. 32. Miller DD, de Ruijter NCA, Bisseling T, Emons AMC: The role of actin in root hair morphogenesis: studies with lipochito-oligosaccharide as a growth stimulator and cytochalasin as an actin perturbing drug. Plant J 1999, 17:141-154. 33. Kost B, Spielhofer P, Chua N-H: A GFP-mouse talin fusion protein labels plant actin ®laments in vivo and visualizes the actin cytoskeleton in growing pollen tubes. Plant J 1998, 16:393-401. 34. Jones MA, Shen J-J, Fu Y, Yang Z, Grierson CS: The Arabidopsis Rop2 GTPase is a positive regulator of both root hair initiation and tip growth. Plant Cell 2002, 14:763-776. The authors combine analysis of ROP2 expression and localization with functional analysis of ROP2 using dominant negative and constitutively active forms of ROP2. They demonstrate a role for this GTPase in the polarization of root hair tip growth. 35. Fu Y, Wu G, Yang Z: Rop GTPase-dependent dynamics of tip-localized F-actin controls tip growth in pollen tubes. J Cell Biol 2001, 152:1019-1032. 36. Lin Y, Wang Y, Zhu J-K, Yang Z: Localization of a Rho GTPase implies a role in tip growth and movement of the generative cell in pollen tubes. Plant Cell 1996, 8:293-303. 37. Li H, Lin Y, Heath RM, Zhu MX, Yang Z: Control of pollen tube tip growth by a Rop GTPase-dependent pathway that leads to tip-localized calcium in¯ux. Plant Cell 1999, 11:1731-1742. 38. Kost B, Lemichez E, Spielhofer P, Hong Y, Tolia K, Carpenter C, Chua N-H: Rac homologs and compartmentalized phosphatidylinositol 4,5-bisphosphate act in a common pathway to regulate polar pollen tube growth. J Cell Biol 1999, 145:317-330. 39. Molendijk AJ, Bischoff F, Rajendrakumar CSV, Friml J, Braun M, Gilroy S, Palme K: Arabidopsis thaliana Rop GTPases are localized to tips of root hairs and control polar growth. EMBO J 2001, 11:2779-2788. 40. Samaj J, Ovecka M, Hlavacka A, Lecourieux F, Meskiene I, Lichtscheidl I, Lenart P, Salaj J, Volkmann D, BoÈgre L et al.: Involvement of the mitogen-activated protein kinase SIMK in regulation of root hair tip growth. EMBO J 2002, 13:3296-3306. Current Opinion in Plant Biology 2003, 6:63±73 45. Vissenberg K, Fry SC, Verbelen J-P: Root hair initiation is coupled to a highly localized increase of xyloglucan endotransglycosylase action in Arabidopsis roots. Plant Physiol 2001, 127:1125-1135. 46. Ringli C, Baumberger N, Diet A, Frey B, Keller B: ACTIN2 is essential for bulge site selection and tip growth during root hair development of Arabidopsis. Plant Physiol 2002, 129:1464-1472. The authors con®rm an important role for actin in the tip growth of root hairs. They also provide important new evidence of a role for actin in the selection and focusing of the root hair bulge site to the apical end of the trichoblast. 47. Quader H, Herth W, Ryser U, Schnepf E: Cytoskeletal elements in cotton seed hair development in vitro: their possible regulatory role in cell wall organization. Protoplasma 1987, 137:56-62. 48. Tiwari SC, Wilkins TA: Cotton (Gossypium hirsutum) seed trichomes expand via diffuse growing mechanism. Can J Bot 1995, 73:746-757. 49. Szymanski DB, Marks MD, Wick SM: Organized F-actin is essential for normal trichome morphogenesis in Arabidopsis. Plant Cell 1999, 11:2331-2347. 50. Mathur J, Spielhofer P, Kost B, Chua N-H: The actin cytoskeleton is required to elaborate and maintain spatial patterning during trichome cell morphogenesis in Arabidopsis thaliana. Development 1999, 126:5559-5568. 51. Kirik V, Grini PE, Mathur J, Klinkhammer E, Adler K, Bechtold N, Herzog M, Bonneville J-M, HuÈlskamp M: The Arabidopsis TUBULIN-FOLDING COFACTOR A gene is involved in the control of the a-/b-tubulin monomer balance. Plant Cell 2002, 14:2265-2276. 52. Kirik V, Mathur J, Grini PE, Klinkhammer I, Adler K, Bechtold N, Herzog M, Bonneville J-M, HuÈlskamp M: Functional analysis of the tubulin folding cofactor C in Arabidopsis thaliana. Curr Biol 2002, 12:1519. 53. Mathur J, Chua N-H: Microtubule stabilization leads to growth reorientation in Arabidopsis trichomes. Plant Cell 2000, 12:465-477. 54. Bouyer D, Kirik D, HuÈlskamp M: Cell polarity in Arabidopsis trichomes. Semin Cell Dev Biol 2001, 12:353-356. 55. Oppenheimer DG, Pollock MA, Vacik J, Szymanski DB, Ericson B, Feldmann K, Marks MD: Essential role of a kinesin-like protein in Arabidopsis trichome morphogenesis. Proc Natl Acad Sci USA 1997, 94:6261-6266. 56. Song H, Golovkin M, Reddy ASN, Endow SA: In vitro motility of AtKCBP, a calmodulin-binding kinesin protein of Arabidopsis. Proc Natl Acad Sci USA 1997, 94:322-327. 57. Krishnakumar S, Oppenheimer DG: Extragenic suppressors of the Arabidopsis zwi-3 mutation identify new genes that www.current-opinion.com Cell shape Smith 73 function in trichome branch formation and pollen tube growth. Development 1999, 126:3079-3088. 58. Bowser J, Reddy ASN: Localization of a kinesin-like calmodulin-binding protein in dividing cells of Arabidopsis and tobacco. Plant J 1997, 12:1429-1437. 59. Vos JW, Safadi F, Reddy ASN, Hepler PK: The kinesin-like calmodulin binding protein is differentially involved in cell division. Plant Cell 2000, 12:979-990. 60. Tsuge T, Tsukaya H, Uchimiya H: Two independent and polarized processes of cell elongation regulate leaf blade expansion in Arabidopsis thaliana (L.) Heynh. Development 1996, 122:1589-1600. 61. Folkers U, Kirik V, SchoÈbinger U, Falk S, Krishnakumar S, Pollock M, Oppenheimer DG, Day I, Reddy AR, JuÈrgens G, HuÈlskamp M: The cell morphogenesis gene ANGUSTIFOLIA encodes a CtBP/ BARS-like protein and is involved in the control of the microtubule cytoskeleton. EMBO J 2002, 21:1280-1288. This study demonstrates that the AN gene plays a role in organizing microtubules in expanding trichomes, that AN encodes a CtBP/BARS-related protein, and that this protein interacts genetically and physically with ZWI. 62. Kim G-T, Shoda K, Tsuge T, Cho K-H, Uchimiya H, Yokoyama R, Nishitani K, Tsukaya H: The ANGUSTIFOLIA gene of Arabidopsis, a plant CtBP gene, regulates leaf-cell expansion, the arrangement of cortical microtubules in leaf cells and expression of a gene involved in cell-wall formation. EMBO J 2002, 21:1267-1279. The AN gene is shown to play a role in organizing microtubules in expanding leaf cells, and to encode a protein that is related to the CtBP/BARS proteins that function in the transcriptional co-repression and regulation of Golgi dynamics. 63. Jung G, Wernicke W: Cell shaping and microtubules in developing mesophyll of wheat (Triticum aestivum L.). Protoplasma 1990, 153:141-148. 64. Panteris E, Apostolakos P, Galatis B: Microtubules and morphogenesis in ordinary epidermal cells of Vigna sinensis leaves. Protoplasma 1993, 174:91-100. 65. Panteris E, Apostolakos P, Galatis B: Sinuous ordinary epidermal cells: behind several patterns of waviness, a common morphogenetic mechanism. New Phytol 1994, 127:771-780. 66. Wasteneys GO, Willingdale-Theune J, Menzel D: Freeze shattering: a simple and effective method for permeabilizing higher plant cell walls. J Microsc 1997, 188:51-61. www.current-opinion.com 67. Frank MJ, Smith LG: A small, novel protein highly conserved in plants and animals promotes the polarized growth and division of maize leaf epidermal cells. Curr Biol 2002, 12:849-853. Analysis of brk1 mutants reveals a previously unsuspected role for local Factin polymerization in the formation of epidermal pavement cell lobes, and suggests that the small, highly conserved BRK1 protein acts to stimulate local F-actin polymerization. In line with this conclusion, the mammalian homolog of BRK1 is has now been directly implicated in the regulation of actin dynamics [72]. 68. Frank MJ, Cartwright HN, Smith LG: Three Brick genes have distinct functions in a common pathway promoting polarized cell growth and division in the maize leaf epidermis. Development 2002, in press. The authors present data that show that two additional genes, Brk2 and Brk3, function in the same pathway as Brk1 to regulate the formation of lobes in epidermal pavement cells. Each brk mutation behaves differently in genetic mosaics, indicating that each Brk gene has a distinct function. 69. Fu Y, Li H, Yang Z: The ROP2 GTPase controls the formation of cortical ®ne F-actin and the early phase of directional cell expansion during Arabidopsis organogenesis. Plant Cell 2002, 14:777-794. Analysis of the function and localization of ROP2 implies key roles for this protein in the regulation of actin polymerization, which is necessary for the shaping of epidermal pavement cells. 70. Qiu J-L, Jilk R, Marks MD, Szymanski DB: The Arabidopsis SPIKE1 gene is required for normal cell shape control and tissue development. Plant Cell 2002, 14:101-118. Analysis of spk1 mutant phenotypes implicates the wildtype SPK1 gene in the cytoskeletal regulation of cell shape in leaf epidermal pavement cells and trichomes. The sequence of SPK1 suggests that it mediates the reorganization of the cytoskeleton in response to extracellular cues, possibly via an interaction with a Rho-related GTPase. 71. Higgs HN, Pollard TD: Regulation of actin ®lament network formation through ARP2/3 complex: activation by a diverse array of proteins. Annu Rev Biochem 2001, 70:649-676. 72. Eden S, Rohtagi R, Podtelejnikov AV, Mann M, Kirschner MW: Mechanism of regulation of WAVE1-induced actin nucleation by Rac1 and Nck. Nature 2002, 418:790-793. A groundbreaking study that elucidates a novel mechanism for the regulation of Arp2,3-dependent actin polymerization. This mechanism involves the Rac/Nck-dependent activation of a complex containing WAVE1 and the mammalian homolog of the maize BRK1 protein, HSPC300. Current Opinion in Plant Biology 2003, 6:63±73