The Growth Cone A Central Player in Axon Pathfinding

The growth cone structure was first characterized more than a century ago by the Spanish neuroana-tomist, Ramon y Cajal. He imagined the growth cone as a ''soft battering ram'' that extending axons used to force their way through the embryonic brain (Ramon y Cajal, 1890). Following this discovery, it was shown that axons of embryonic neural tube tissue are tipped with growth cones and are able to grow along a glass coverslip (Harrison, 1910). These very active structures have now been characterized using imaging techniques in both invertebrates and vertebrates and their structure shows a remarkable degree of conservation among species (reviewed in Sanes et al., 2000).

1.12.2.1 The Growth Cone is a Highly Dynamic Structure and Changes Its Shape in Response to Its Environment

1.12.2.1.1 Structure and organization of the growth cone The growth cone is a highly dynamic, actin-rich structure capable of recognizing and responding to a variety of guidance cues (Figure 1). It consists of two major domains, the peripheral and central domains, which are characterized by specific cytoskeletal components. The peripheral domain is rich in actin filaments, while the central domain mainly contains microtubules, mitochondria, and various other organelles. In the peripheral domain, the actin filaments are organized into different types of membrane structure: they can extend into finger-like protrusions (filopodia), remain closely associated with the substratum (lamellipodia), or appear at the apical surface of the growth cone (membrane ruffles).

Microtubules present in the central domain are oriented with their plus ends toward or within the peripheral domain (Gordon-Weeks, 1993) and serve as the major cytoskeletal structure on which rapid vesicular transport occurs (Lockerbie et al., 1991 Such vesicles deliver the molecules needed for neurite elongation, including structural proteins and lipids.

Peripheral domain

F-actin bundles

/ Filopodium

Peripheral domain

F-actin bundles

/ Filopodium

Microtubule filament

Central domain

Microtubule bundles

Figure 1 Structure of the growth cone. The growth cone is the highly dynamic tip of the axon that is involved in pathfinding. The growth cone consists of a central and a peripheral domain. The central domain contains stable microtubule bundles as well as membranous organelles. The peripheral domain is rich in actin filaments, which comprise either the F-actin network found in lamellipodia, or the F-actin bundles of filopodia. The growth cone cytoskeleton can be rapidly remodeled in response to environmental cues. Adapted from Dickson, B. J. 2002. Molecular mechanisms of axon guidance. Science 298, 1959-1964.

Microtubule bundles

F-actin network in lamellipodium

Microtubule filament

Central domain

Figure 1 Structure of the growth cone. The growth cone is the highly dynamic tip of the axon that is involved in pathfinding. The growth cone consists of a central and a peripheral domain. The central domain contains stable microtubule bundles as well as membranous organelles. The peripheral domain is rich in actin filaments, which comprise either the F-actin network found in lamellipodia, or the F-actin bundles of filopodia. The growth cone cytoskeleton can be rapidly remodeled in response to environmental cues. Adapted from Dickson, B. J. 2002. Molecular mechanisms of axon guidance. Science 298, 1959-1964.

1.12.2.1.2 The dynamic cytoskeleton The growth cone cytoskeleton plays an important role in the process of cell movement (Reinsch et al., 1991; Bentley and O'Connor, 1994; Heidemann et al., 1995; Heidemann, 1996; Letourneau, 1996). The most important cytoskeleton components for axon pathfinding are actin, tubulin, and several actin- and microtubule-associated proteins (MAPs). Tubulin and actin polymerize at the distal tip to generate microtubules and microfilaments respectively. Associated proteins are involved in the assembly, disassembly, and stabilization of actin and tubulin as well as in the anchoring of actin and microtubules to the cell membrane or to other cytoskeletal components. For example, myosin, an actin-associated protein, is able to generate a vectorial force in the growth cone by pulling on the actin filaments, a mechanism similar to muscle contraction.

The filopodia dynamics largely account for the sensory capacity of the growth cone. The length of the filopodia (tens of microns, in some cases) allows them to search within the environment and to navigate across cells and obstacles. Movement at the tips of growing filopodia is generated by the rapid assembly of actin filaments, whereas microtubule assembly is involved in the advance of the body of the growth cone. Importantly, experiments performed in vitro have demonstrated that a single filopodium making strong contact with an adhesive substrate is able to steer the growth cone by pulling it toward the adhesive substrate (Letourneau, 1996). Furthermore, when one filopodium is detached, the growth cone changes direction due to the release of tension from that side (Wessells, 1978).

Many of our insights into the role of cytoskeleton in axon pathfinding come from experiments using depolymerizing drugs that interfere with the normal function of microtubules and filopodia. For example, cytochalasin B, an actin-depolymerizing drug, prevents filopodia formation, resulting in axons either stopping growing or slowing dramatically (Bentley and Toroian-Raymond, 1986). Moreover, retinal axons of the amphibian brain treated with cytochalasin grow past a critical turning point and fail to find their targets (Chien et al., 1993). Indeed, experiments using the tibial (Ti1) pioneer neuron in the grasshopper limb have shown that cytochalasin treatment of growth cones induces a loss of filopo-dia, but the lamellipodia are still functional, allowing the growth cone to advance slowly (Bentley and Toroian-Raymond, 1986). The role of filopodia in directing growth cones has been demonstrated in the pioneer axons of the grasshopper limb. When a single filopodium of a Ti1 neuron makes contact with a guidepost cell indicating the direction to choose, then this filopodium stays in contact while other filopodia retract. The filopodium in contact with the guidepost cell is stabilized and eventually becomes the shaft of the growing axon (Sabry et al., 1991). These results indicate that filo-podia play a critical role in axon growth and navigation during pathfinding.

Interfering with microtubule dynamics with pharmacological inhibitors has also provided evidence highlighting their role in axon growth. It has been shown that axons lacking filopodia by treatment with cytochalasin are still able to grow due to the addition of tubulin at the plus end of the microtubule at the distal process. However, axon elongation can be completely inhibited by the depolymerization of microtubules at the distal tip, the most sensitive region of the growing axon to drugs (Marsh and Letourneau, 1984; Bentley and Toroian-Raymond, 1986).

In addition to microtubules, axonal growth also involves several MAPs. For example, treatment of cultured neurons with the nerve growth factor (NGF), a growth factor that promotes neurite outgrowth, results in the upregulation of the MAPs Tau and MAP1B, and inhibition of Tau function can block neurite outgrowth, indicating the critical role of this protein (Letourneau, 1996). As the coupling between the sensory and motor capabilities of the growth cone is critical for axon pathfinding, it has been hypothesized that actin and microtubules may be associated to generate cell movement (Letourneau, 1996; Williamson et al., 1996). Indeed, some MAPs may also bind actin, providing a mechanical link between these two structural components. However, the mechanism of growth cone steering and how these cytoskeletal components are regulated by the environmental milieu are still incompletely understood.

1.12.2.1.3 Growth cones change their morphologies in response to the environment En route to their target, growth cones advance within their environment and display a range of morphologies, from simple cigar shapes to highly complex structures, depending on their position (Tosney and Landmesser, 1985; Bovolenta and Mason, 1987). When axons are growing in fasciculated nerve bundles along straight tracts, they often display a streamlined form with one or two filopodia pointing in the direction of growth. Studies in mice and in Xenopus have shown that growth speed is about 55mmh-1 (Harris et al., 1987; reviewed in Mason and Erskine, 2000). Growth is, however, often saltatory, meaning that rapid advance is frequently interrupted by pauses. These breaks are mainly observed at critical choice points, where decisions regarding subsequent directional growth are made.

At choice points, where growth cones display highly complex morphologies, they bear a widespread lamellipodium and multiple filopodia. Indeed, it has been shown that at key choice points in the retinal pathway (see Section 1.12.3), growth cones pause for periods of approximately 30min to determine which direction to follow (Llirbat and Godement, 1999). These morphological changes indicate that the growth cone needs to integrate signals from the environment to advance along its pathway.

1.12.2.2 Axon-Pathfinding Concepts

We have seen that the growth cone serves as both an antenna that receives directional cues from the environment and as a motor structure that drives axon growth. How do axons find their target in the highly complex environment of the developing CNS? To visualize this complex process, axon path-finding can be compared to a maze (Figure 2). During its long journey, the axon will have to choose which direction to take by interpreting attractive or repulsive signals from the environment at each decision point. To navigate through the maze of the developing embryo to reach their final destination, growing axons are guided by various diffusible or substrate-bound molecules, which they use as navigational cues. Interestingly, the decisions at choice points are not made by trial and error; instead, the growth cone pauses, samples the environment with filopodia, and adapts its direction accordingly, resulting in an almost error-free navigation. Importantly, when growth cones are cut from their cell bodies, they still continue to navigate appropriately in the brain and detect cues, demonstrating that growth cones can function independently of their cell bodies (Harris et al., 1987).

Our understanding of axon pathfinding comes from a variety of experimental data generated in both invertebrate and vertebrate model systems. The nervous system architecture of invertebrates is relatively simple, which allows researchers to trace individual neurons, so providing a useful tool to perform in vivo experiments. Remarkably, axon-pathfinding concepts are very well conserved and have also been found in higher organisms.

An illustrative example of the concept of axon pathfinding is the experiment performed by Sperry in 1943. In this experiment, he severed the optic nerve in a frog and rotated the eye by 180° (Sperry, 1944). After regeneration and restoration of the retinotectal projection, the animal behaved as if it saw the world upside down and back to front. Therefore, Sperry proposed his chemospecificity hypothesis, which stated that each individual neuron had its specific chemical tag to find its target. Twenty years later he refined his theory and proposed that guidance cues are expressed in gradients, which would be more economical, requiring fewer molecules than individual labels would. Gradients would also have the advantage of being able to tell the axon the direction to go, removing the necessity of searching for the target randomly. Supporting this theory, rotation experiments performed in salamander hind-brains revealed that axons still find their way to their targets, suggesting that external cues are involved in this pathfinding rather than molecules secreted by the cell bodies (Hibbard, 1965).

Complementary evidence that growth cones are guided by external cues comes from experiments performed in vivo using the grasshopper leg (Keshishian and Bentley, 1983; Raper et al., 1983). During development of the grasshopper, a pair of afferent neurons arises in the distal tip of each limb bud. These Ti1 pioneer neurons navigate to their

Cell body

Cell body

Cell

Figure 2 Axon pathfinding. The environment through which a growing axon must traverse to reach its target is analogous to a maze. The axon negotiates the maze of the developing nervous system by responding to specific positive or negative directional cues at intermediate choice points along its pathway.

final destination following a stereotypical pathway. Experiments using two sibling neurons have shown that growth cones of later-differentiating neurons elongate upon the axons of specifically labeled earlier-differentiating neurons (Raper et al., 1983). When pioneer axons were ablated with a laser, the later axons could not find their way into the CNS (Keshishian and Bentley, 1983). To find their correct target, the pioneer Ti1 axons are guided by local cues or guidepost cells that are spaced at short distances (Bentley and Caudy, 1983). When these cells were obliterated using a laser, the Ti1 axons were unable to travel from one segment to another (Bentley and Caudy, 1983). In addition to these critical guidepost cells, adhesive molecules on the epithelium are also involved in directing the growth cone from one guidepost cell to the next and growth-inhibitory molecules prevent the growing axons from taking the wrong way (Singer et al., 1995).

Was this article helpful?

0 0

Post a comment