Locomotion clearly sets plants and animals apart. However, recent studies in higher plants reveal cell-biological and molecular features similar to those observed at the leading edge of animal cells and suggest conservation of boundary extension mechanisms between motile animal cells and nonmotile plant cells.
Boundary extension: a necessity for living organisms
Survival requires a constant refurbishment of resources. Therefore, living organisms persistently explore their environment for extending their boundaries into favorable regions. One fundamental difference we are taught early on between motile animal and nonmotile plant cells is the way in which they carry out their boundary extension. Motile cells achieve whole-body displacement or locomotion, which allows them to leave resource-depleted domains behind while moving on to new areas. Plant cells, however, become rooted to a spot and are only able to extend outwards from it in response to different stimuli.
Two major conditions have to be met for achieving locomotion; a flexible bounding layer of the cell, which allows for rapid change in its shape, and the ability to retract one end of the cell while extending the other (Small et al., 2002a). Plant cells do not fulfill these requirements because they are encased in a relatively rigid cellulosic cell wall and, clearly, they do not have rear-end retractability. However, the cell wall of actively expanding plant cells is actually quite labile in contrast to its later, rigid nature in a mature, fully expanded cell (Mathur, 2004). Further, though lacking rear-end retractability, the forward extension of a plant cell, like that of an animal cell, also involves membrane protrusion (Vidali et al., 2001).
Recent cell-biological and molecular evidence from higher plants suggests that, despite their sessile and wall-encased nature, plant cells possess a core machinery similar to the one required for forward motility of animate cells.
Boundary extension requires a leading edge
Although usually used in the context of crawling amoeboid locomotion, here I use the term “leading edge” for both animal and plant cells as the part of a cell that extends or “leads.”
Though descriptions vary from cell to cell, the generalized leading edge of an amoeboid cell comprises a 2–5-μm-wide veil-like, organelle free, cytoplasmic extension called the lamellipodium. The leading edge has an actin-rich zone comprising of a fine F-actin mesh followed by a microtubule-rich region (Etienne-Manneville, 2004) (Fig. 1 A). A few pioneering microtubules do extend into the F-actin mesh (Small et al., 2002b; Raftopoulou and Hall, 2004).
A very similar intracellular zonation is seen in plants (Fig. 1, B–D; Mathur, 2004). For plant cells, two kinds of expansion modes are recognized; one called tip-growth, where the growth processes is limited to a small region that extends to form a tubular structure, and the second, designated diffuse growth, where the process of growth is dispersed over a large area of the cell (Mathur, 2004). Tip-growing cells (Fig. 1, B and D) are best typified by elongating root-hair and pollen tubes and, like lamellipodia, exhibit an apical region with a stretched plasma membrane, followed by a organelle-depleted clear zone (Vidali and Hepler, 2001). A fine, labile F-actin zone is defined next and leads into a dense F-actin region where the filaments become progressively bundled (Fu et al., 2001; Ketelaar et al., 2003). In active tip-growing cells, most cytoplasmic microtubules extend only to the edge of the fine F-actin mesh (Sieberer et al., 2002). A large vacuole fills the rest of the cell (Carol and Dolan, 2002). By comparison, the volume occupied by vacuoles is quite large in diffuse-growing cells so that the cytoplasm is compressed into a thin layer against the plasma membrane (Fig. 1, B and C; Mathur et al., 2003a,b). Although this thin cytoplasmic layer has not allowed intracellular zonation in these cells to be appreciated as clearly as in tip-growing cells, a fine cortical mesh, underlying cytoplasmic F-actin bundles (Fu et al., 2002; Mathur et al., 2003b), and microtubules (Saedler et al., 2004a) observed in different diffuse-growing cells suggest a very similar zonation (Fig. 1 C). The apparent similarity in zonal relationship between the plasma membrane and cytoskeletal elements in the region of active protrusive growth between animal and plant cells (Fig. 1) extends to the molecular mechanisms responsible for the creation of a leading edge.
Creating the leading edge
A considerable body of information exists about the multiple protein complexes and regulatory molecules that get activated to create the leading edge in an animal cell (Kraynov et al., 2000; Pollard and Borisy, 2003; Rodriguez et al., 2003). Three key interacting components—members of the Rho/rac/Cdc42 superfamily of GTPases, and actin and microtubule cytoskeletons—stand out. Although Rho-GTPase loops interact with and regulate both actin and microtubule cytoskeletons and microtubules provide directionality to the leading edge, it is actin dynamics that play a major role in membrane protrusion (Small et al., 2002b; Pollard and Borisy, 2003). One of the proposed pathways for actin–cytoskeleton regulation in animal cells involves Cdc42/Rac-GTPase–triggered activation of a suppressor of cAMP receptor from Dictyostelium (SCAR)/Wiskott-Aldrich syndrome protein family verprolin-homologous protein (WAVE) complex. This complex in turn activates a 7-subunit actin-related protein (ARP2/3) complex to enhance actin polymerization (Millard et al., 2004; Vartiainen and Machesky, 2004). The final outcome is a fine dendritic mesh of filamentous actin seen at the leading edge (Fig. 1).
Plants, too, possess a unique subfamily of Rho-family GTPases, called Rho-related GTPase of plants (ROP; Vernoud et al., 2003), whose members localize to areas of active growth (Fu et al., 2001, 2002; Fig. 1, C and D) and play numerous roles, including that of actin–cytoskeleton regulation (Gu et al., 2004). Further, different components of a putative SCAR/WAVE-like complex (Millard et al., 2004; Vartiainen and Machesky, 2004), namely SCAR-related proteins (Frank et al., 2004), NAP125/GNARLED (Brembu et al., 2004; Deeks et al., 2004; El-Din El-Assal et al., 2004; Li et al., 2004; Zimmermann et al., 2004), PIR121/KLUNKER/PIROGI (Basu et al., 2004; Brembu et al., 2004; Li et al., 2004; Saedler et al., 2004b), HSPC300/BRICK1 (Frank and Smith, 2002), and Ab1-1–like proteins (Deeks et al., 2004), have been cloned from Arabidopsis. At least one ROP, AtROP2, has been shown to interact with PIR121, strongly suggesting that a SCAR/WAVE-like pathway mediates the ROP–actin interaction in plants (Basu et al., 2004). Finally, homologues of the different subunit of the ARP2/3 complex have been cloned from plants and in some cases shown to be interchangeable with their animal orthologues (Le et al., 2003; Li et al., 2003; Mathur et al., 2003a,b; El-Assal et al., 2004; Saedler et al., 2004a).
In animal cells, changes in actin-mesh density are believed to affect microtubule plus-end growth by removing a stearic hindrance that allows cytoplasmic microtubules localized access to cortical domains (Rodriguez et al., 2003). In plants, studies on actin–microtubule interactions are just beginning (Mathur, 2004). However, observations of an aberrant aggregation and stabilization of cytoplasmic microtubules in actin-compromised cells suggests an actin control over microtubules very similar to that observed in animal cells (Saedler et al., 2004a). Alternatively, microtubule control over the actin cytoskeleton has also been demonstrated in Drosophila by providing evidence that EB1, a microtubule plus end–binding protein, mediates the delivery of DRhoGEF2, an activator of Rho1 to cortical domains for its actomyosin-related functions in cell retraction (Rogers et al., 2004). Though not yet demonstrated, similar microtubule–ROP interactions could be envisaged in plants too, as several microtubule plus end–binding proteins (Bisgrove et al., 2004) as well as a putative ROP-GEF, SPIKE1, have been cloned from Arabidopsis. Interestingly, cells of the spike1 mutant display an aberrant microtubule cytoskeleton (Qiu et al., 2002).
Defects at the leading edge: mutant phenotypes
Compromised function of the three key molecular components, as mentioned in the previous section, has major implications for both motility and cell shape. A brief comparison of cellular phenotypes resulting from defects in the core elements underscores the commonalities. Actin-dependent lamellipodium protrusion and cell motility get greatly attenuated in animal cells with an aberrant actin cytoskeleton (Etienne-Manneville, 2004). Likewise, cell expansion is significantly reduced in actin-compromised plant cells (Baluska et al., 2001). This is especially apparent in Arabidopsis mutants such as nap125/gnarled, pir121/klunker/pirogi (Brembu et al., 2004; Deeks et al., 2004; El-Assal et al., 2004; Li et al., 2004; Saedler et al., 2004b; Zimmermann et al., 2004), and hspc300/brick1 (mutant in maize; Frank and Smith, 2002), whose respective gene products possibly feed into the ARP2/3 complex regulatory pathway. Mutations in different subunit homologues of the putative plant ARP2/3 complex such as arp2/wurm, arp3/distorted1, arpc2/distorted2, and arpc5/crooked also exhibit similar phenotypes (for review see Mathur, 2005). Each of these mutants exhibits aberrant F-actin organization and characteristic cellular deformations resulting from abnormal, random expansion. However, as mentioned earlier in this paper, actin cytoskeleton activity is heavily dependent upon the stimulation provided by Rho-GTPases. Animal cells with aberrant Rho-GTPase activity display numerous defects in pseudopod extension and chemotaxis (Chung et al., 2000; Raftopoulou and Hall, 2004). Overexpression of ROPs in plants results in increased fine F-actin meshworks and a swollen cell morphology, suggesting an overall reduction in growth polarization (Molendijk et al., 2001; Fu et al., 2002; Jones et al., 2002). Because loss of polarity is a characteristic feature of microtubule defective cells too, in both animal (Small et al., 2002b) and plant cells (Mathur, 2004) these observations again point to a conserved, intimate relationship between Rho-family GTPases, and the actin and microtubule cytoskeletons.
Although the similarities described above emphasize conserved mechanisms for membrane protrusion between animal and plant cells, they also reflect the profound impact the mechanism of boundary extension has on the survival ability of an organism. For motile cells, loss of motility leads to an inability to move to resource-rich areas and ultimately leads to death. Consequently, many animal mutants for the core elements display lethal phenotypes. For the nonmotile but photosynthetic (and therefore relatively self-sufficient) plant cells, a change in cell shape does not lead to death directly. As a result, many comparable plant mutants, though misshapen and underweight (El-Assal et al., 2004; Li et al., 2004), are able to grow and complete their life cycle, giving the erroneous impression that the core genes are not as vital for plants as for animals. However, when grown in a population of siblings, the inability of mutant plants to extend properly becomes a severe handicap that affects their survival ability.
Conclusions and perspectives
A comparison of key cell biological and molecular features in two apparently disparate processes, namely, forward motility of animal cells and extension growth of plant cells, reveals a core machinery that has been conserved by both motile and nonmotile life forms for achieving boundary extension; an essential requirement for survival. Given this realization, future studies on boundary extension mechanisms of different life forms should prove exciting, not only for uncovering the extent of similarities, but also for discovering the range of cell biological and molecular variations, and the adaptations and innovations introduced by different organisms for surviving in their specialized niches.
Abbreviations used in this paper: ARP, actin-related protein; ROP, Rho-related GTPase of plants; SCAR, suppressor of cAMP receptor; WAVE, Wiskott-Aldrich syndrome protein family verprolin-homologous protein.