Cellular Strategies of Axonal Pathfinding

Axons follow highly stereotyped and reproducible trajectories to their targets. In this review we address the properties of the first pioneer neurons to grow in the developing nervous system and what has been learned over the past several decades about the extracellular and cell surface substrata on which axons grow. We then discuss the types of guidance cues and their receptors that influence axon extension, what determines where cues are expressed, and how axons respond to the cues they encounter in their environment.This article provides an overview of how growth cones respond to the cellular substrata and molecular cues they encounter as they extend through the developing nervous system. It elaborates on the primer by Kolodkin and Tessier-Lavigne (2010) and touches on many of the topics covered in greater detail in the articles that follow. The first sections describe how axons extend in a directed manner, the substrata on which they grow, interactions between pioneer and follower axons, and growth cone behaviors in emerging tracts and at decision points. The subsequent sections discuss examples of specific cues, their distributions, how their distributions are determined, and how growth cones integrate multiple cues during pathfinding.     

Growth cones stall and collapse during axon outgrowth in Caenorhabditis elegans.

During nervous system development, neurons form synaptic contacts with distant target cells. These connections are formed by the extension of axonal processes along predetermined pathways. Axon outgrowth is directed by growth cones located at the tips of these neuronal processes. Although the behavior of growth cones has been well-characterized in vitro, it is difficult to observe growth cones in vivo. We have observed motor neuron growth cones migrating in living Caenorhabditis elegans larvae using time-lapse confocal microscopy. Specifically, we observed the VD motor neurons extend axons from the ventral to dorsal nerve cord during the L2 stage. The growth cones of these neurons are round and migrate rapidly across the epidermis if they are unobstructed. When they contact axons of the lateral nerve fascicles, growth cones stall and spread out along the fascicle to form anvil-shaped structures. After pausing for a few minutes, they extend lamellipodia beyond the fascicle and resume migration toward the dorsal nerve cord. Growth cones stall again when they contact the body wall muscles. These muscles are tightly attached to the epidermis by narrowly spaced circumferential attachment structures. Stalled growth cones extend fingers dorsally between these hypodermal attachment structures. When a single finger has projected through the body wall muscle quadrant, the growth cone located on the ventral side of the muscle collapses and a new growth cone forms at the dorsal tip of the predominating finger. Thus, we observe that complete growth cone collapse occurs in vivo and not just in culture assays. In contrast to studies indicating that collapse occurs upon contact with repulsive substrata, collapse of the VD growth cones may result from an intrinsic signal that serves to maintain growth cone primacy and conserve cellular material.     

Multiple strategies for directed growth cone extension and navigation of peripheral neurons

Leeches have a diverse constellation of peripheral neural elements that are challenged to extend growth cones in highly specific ways in a constantly changing embryonic environment. Two major systems are reviewed here. In one, peripheral afferents extend growth cones toward the central nervous system (CNS), forming common pathways, and then segregate into particular tracts within the CNS. A majority of these afferents depend on CNS-derived guidance cues and projections from the CNS to guide their way. However, not all of the nerves are established this way and at least one of the peripheral nerves is likely to be pioneered by sensillar sensory afferents. The distribution of particular antigens (such as the lan3–2 antigen) suggests the identity of molecules involved in homophilic adhesion along common pathways, whereas others (such as the lan4–2 and 3–6 antigens) are candidates for mediating specific pathway choices. In the second system, the myo-organizing Comb cell (C cell) projects multiple growth cones simultaneously along oblique trajectories not influenced by segmental or midline boundaries. Its parallel growth cones exhibit space-filling as well as directional growth and are guided by local cues to extend in discrete phases that are coordinated with the development of the environment. Both systems exhibit highly directed outgrowth orchestrated by a hierarchy of cues, establish patterns of neurites used to direct later migrating cells, and seem to be regulated temporally and spatially by interactions with the embryonic environment. These systems illustrate the strengths of examining neural development in vivo across several levels of analysis. © 1995 John Wiley & Sons, Inc.     

Immunolocalization of a neuronal growth-dependent membrane glycoprotein

Monoclonal antibody (mAb) 5B4 recognizes in the rat a large, developmentally regulated membrane glycoprotein. The larger form of this antigen (185-255 kD) occurs in the developing nervous system and is present in membranes of nerve growth cones, as determined by analysis of a growth cone particle fraction. An immunochemical characterization of this antigen and of a smaller form (140 kD), sparsely present in the mature nervous system, has been described (Ellis, L., I. Wallis, E. Abreu, and K. H. Pfenninger, 1985, J. Cell. Biol., 101:1977-1989). The present paper reports on the localization by immunofluorescence of 5B4 antigen in cultured cortical neurons, developing spinal cord, and the mature olfactory system. In culture, mAb 5B4 stains only neurons; it is sparsely present in neurons at the onset of sprouting while, during sprouting, it appears to be concentrated at the growth cone and in regions of the perikaryon. In the developing spinal cord, 5B4 labeling is faintly detectable on embryonic day 11 but is intense on fetal day 13. At this stage, the fluorescence is observed in regions of the cord where axonal growth is occurring, while areas composed of dividing or migrating neural cells are nonfluorescent. With maturation of the spinal cord, this basic pattern of fluorescence persists initially, but the staining intensity decreases dramatically. In the adult, faint fluorescence is detectable only in gray matter, presumably indicating the presence of the 140 kD rather than the fetal antigen. The only known structure of the adult mammalian nervous system where axonal growth normally occurs is the olfactory nerve. mAb 5B4 intensely stains a variable proportion of olfactory axons in the mucosa as well as in the olfactory bulb. Based on both immunochemical and immunofluorescence data, the 5B4 antigen of 185-255 kD is associated specifically with growing neurons, i.e., neurons that are generating neurites.     

In vitro studies on the control of nerve fiber growth by the extracellular matrix of the nervous system

1. Cultured neurons from embryonic chick sympathetic ganglia or dorsal root ganglia grow nerve fibers extensively on simple substrata containing fibronectin, collagens (types I, III, IV), and especially laminin. 2. The same neurons cultured on substrata containing glycosaminoglycans grow poorly. Glycosaminoglycans (heparin) inhibit nerve fiber growth on fibronectin substrata. 3. Proteolytic fragments of fibronectin support nerve fiber growth only when the cell attachment region is intact. For example, a 105 kD fragment, encompassing the cell attachment region, supports growth when immobilized in a substratum, but a 93 kD subfragment, lacking the cell attachment region, is unable to support fiber growth. When it is added to the culture medium, the 105 kD fragment inhibits fiber growth on substrata containing native fibronectin. 4. In culture medium lacking NGF, DRG neurons extend nerve fibers only on laminin and not on fibronectin, collagen or polylysine. Studies with radioiodinated laminin indicate that laminin binds with a relatively high affinity (kd approximately equal to 10(-9) M) to DRG neurons, and to a variety of other neural cells (NG108 cells, PC12 cells, rat astrocytes, chick optic lobe cells). We have isolated a membrane protein (67 kD) by affinity chromatography on laminin columns and are characterizing this putative laminin receptor. 5. Dissociated DRG neurons or ganglionic explants cultured on complex substrata consisting of tissue sections of CNS or PNS tissues extend nerve fibers onto the PNS (adult rat sciatic nerve) but not CNS (adult rat optic nerve) substrata. Other tissue substrata which support fiber growth in vivo (embryonic rat spinal cord, goldfish optic nerve) support growth in culture. While substrata from adult CNS, which support meager regeneration in vivo (adult rat spinal cord) support little fiber growth in culture. 6. Ganglionic explants cultured in a narrow space between a section of rat sciatic nerve and optic nerve grow preferentially onto the sciatic nerve suggesting that diffusible growth factors are not responsible for the differential growth on the two types of tissues. 7. Dissociated neurons adhere better to sections of sciatic nerve than optic nerve. Laminin, rather than fibronectin or heparan sulfate proteoglycan, is most consistently identifiable by immunocytochemistry in tissues (sciatic nerve, embryonic spinal cord, goldfish optic nerve) which support nerve fiber growth. Taken together, these data suggest that ECM adhesive proteins are important determinants of nerve regeneration.     

Pioneer growth cone behavior at a differentiating limb segment boundary in the grasshopper embryo

The first neurons to extend axons through embryonic grasshopper limbs are a pair of sibling pioneer neurons. After migrating proximally along the limb axis, the pioneer growth cones normally make an abrupt ventral turn. In some cases (less than 20%) this turn is directly toward the proximo-ventrally located Cx1 guidepost neurons. However, in the majority of cases (greater than 80%) the pioneer growth cones make a more acute ventral turn along a single circumferential line which lies distal to the Cx1 neurons. Growth cones from other afferent neurons orient along the same line. Growth cones can extend along this line around more than half of the circumference of the limb and can grow in either direction along it. The circumferential line appears to be the prospective trochanter-coxa segment boundary. Afferent axons on the segment boundary leave it and contact the proximo-ventrally located Cx1 neurons. The site at which pioneer growth cones leave the boundary is variable and appears to be the point from which filopodial contact with Cx1 cells is first established. In addition to the trochanter-coxa segment boundary, the pioneer growth cones and axons also respond to the tibia-femur and femur-trochanter segment boundaries. The role of segment boundaries as barriers to growth cone movement and the effect of such barriers on the timing and placement of differentiation of pioneer neurons are discussed.     

The role of repulsive guidance molecules in the embryonic and adult vertebrate central nervous system

During the development of the nervous system, outgrowing axons often have to travel long distances to reach their target neurons. In this process, outgrowing neurites tipped with motile growth cones rely on guidance cues present in their local environment. These cues are detected by specific receptors expressed on growth cones and neurites and influence the trajectory of the growing fibres. Neurite growth, guidance, target innervation and synapse formation and maturation are the processes that occur predominantly but not exclusively during embryonic or early post-natal development in vertebrates. As a result, a functional neural network is established, which is usually remarkably stable. However, the stability of the neural network in higher vertebrates comes at an expensive price, i.e. the loss of any significant ability to regenerate injured or damaged neuronal connections in their central nervous system (CNS). Most importantly, neurite growth inhibitors prevent any regenerative growth of injured nerve fibres. Some of these inhibitors are associated with CNS myelin, others are found at the lesion site and in the scar tissue. Traumatic injuries in brain and spinal cord of mammals induce upregulation of embryonic inhibitory or repulsive guidance cues and their receptors on the neurites. An example for embryonic repulsive directional cues re-expressed at lesion sites in both the rat and human CNS is provided with repulsive guidance molecules, a new family of directional guidance cues.     

Promoting and directing axon outgrowth

Establishment of appropriate neuronal connections during development and regeneration requires the extension of processes that must then grow in the correct direction, find and recognize their targets, and make synapses with them. During development, embryonic neurons gradually establish central and peripheral connections in an evolving cellular environment in which neurotrophic factors are provided by supporting and target cells that promote neuronal survival, differentiation, and process outgrowth. Some cells also release neurotropic factors that direct the outgrowth of neuronal processes toward their targets. Following development the neurotrophic requirements of some adult neurons change so that, although they respond to neurotrophic factors, they no longer require exogenous neurotrophins to survive or to extend processes. Within the central nervous system (CNS), the ability of neurons to extend processes is eventually lost because of a change in their cellular environment from outgrowth permissive to inhibitory. Thus, neuronal connections that are lost in the adult CNS are rarely reestablished. In contrast, the environment of the adult peripheral nervous system fosters process outgrowth and synapse formation. This article discusses the neurotrophic requirements of embryonic and adult neurons, as well as the importance of neurotropic factors in directing the outgrowth of regenerating adult axons.     

A specific brain tract guides follower growth cones in two regions of the zebrafish brain.

Neurons of the nucleus of the posterior commissure (nuc PC), an identifiable cluster of neurons in the embryonic zebrafish brain, project growth cones ventrally along the posterior commissure to the anterior tegmentum where the PC intersects two longitudinal tracts, the tract of the postoptic commissure (TPOC) and the medial longitudinal fasciculus (MLF). Once at the intersection, nuc PC growth cones turn posteriorly onto the TPOC in the dorsal tegmentum and follow it to the hindbrain. Previously we showed that in the absence of the TPOC, nuc PC growth cones often extended along aberrant pathways suggesting that fasciculation, that is, contact with TPOC axons is an important factor in guiding growth cones along their normal pathway. However, a significant number of nuc PC growth cones also followed their normal pathway suggesting that cues associated with the dorsolateral tegmentum, independent of the TPOC, can also guide nuc PC growth cones. We have now confirmed using electron microscopy that nuc PC growth cones fasciculate with axons in the TPOC. In the absence of the TPOC, the nuc PC growth cones that extend along their normal pathway do so in contact with dorsolateral neuroepithelial cells. This suggests that cues associated with these cells can also guide the nuc PC growth cones. Furthermore, in the absence of the TPOC axons, these growth cones now inappropriately turn onto axons that normally intersect the TPOC near the border of the midbrain and hindbrain, that is, at a second intersection of tracts. This suggests that fasciculation with TPOC axons may also guide nuc PC growth cones in this second region of the brain.     

A specific brain tract guides follower growth cones in two regions of the zebrafish brain

Neurons of the nucleus of the posterior commissure (nuc PC), an identifiable cluster of neurons in the embryonic zebrafish brain, project growth cones ventrally along the posterior commissure to the anterior tegmentum where the PC intersects two longitudinal tracts, the tract of the postoptic commissure (TPOC) and the medial longitudinal fasciculus (MLF). Once at the intersection, nuc PC growth cones turn posteriorly onto the TPOC in the dorsal tegmentum and follow it to the hindbrain. Previously we showed that in the absence of the TPOC, nuc PC growth cones often extended along aberrant path ways suggesting that fasciculation, that is, contact with TPOC axons is an important factor in guiding growth cones along their normal pathway. However, a significant number of nuc PC growth cones also followed their normal pathway suggesting that cues associated with the dorsolateral tegmentum, independent of the TPOC, can also guide nuc PC growth cones. We have now confirmed using electron microscopy that nuc PC growth cones fasciculate with axons in the TPOC. In the absence of the TPOC, the nuc PC growth cones that extend along their normal pathway do so in contact with dorsolateral neuroepithelial cells. This suggests that cues associated with these cells can also guide the nuc PC growth cones. Furthermore, in the absence of the TPOC axons, these growth cones now inappropriately turn onto axons that normally intersect the TPOC near the border of the midbrain and hindbrain, that is, at a second intersection of tracts. This suggests that fasciculation with TPOC axons may also guide nuc PC growth cones in this second region of the brain. © 1992 John Wiley & Sons, Inc.     

A multifunctional cell surface developmental stage-specific antigen in the cockroach embryo: involvement in pathfinding by CNS pioneer axons

mAb DSS-8 binds to a 164-kD developmental stage-specific cell surface antigen in the nervous system of the cockroach, Periplaneta americana. The antigen is localized to different subsets of cells at various stages of development. The spatial and temporal distributions of DSS-8 binding were determined and are consistent with this antigen playing multiple roles in the development of the nervous system. Direct identification of some of these functions was made by perturbation experiments in which pioneer axon growth occurs in embryos that are cultured in vitro in the presence of mAb DSS-8 or its Fab fragment. Under these conditions the pioneer axons of the median fiber tract grow but follow altered pathways. In a smaller percentage of the ganglia, the immunoreagents additionally produce defasciculation of a subset of DSS-8 labeled axons. Therefore, direct roles for the DSS-8 antigen in both the guidance of pioneer axons and selective fasciculation have been demonstrated.     

Monoclonal antibody 2G13, a new axonal growth cone marker

Growth cones are specialized sensorimotor structures at the tips of neurites implicated in pathfinding decisions and axonal outgrowth during neuronal development. We generated a mouse monoclonal antibody (mAb 2G13) against chick tectum and found that the antibody exclusively labelled axonal growth cones, particularly their filopodia and lamellipodia, in developing rat CNS and in embryonic neurons in culture. The high fidelity of the staining of growth cones by mAb 2G13 means that the antibody will be a useful marker for identifying growth cones. In growth cones of cultured neurons, mAb 2G13 labelling is intracellular and mainly associated with the filamentous actin cytoskeleton. Experiments with cytochalasins, which depolymerise filamentous actin, showed that 2G13p (the protein recognised by mAb 2G13) is physically associated with filamentous actin in growth cones. These properties of 2G13p suggest a role in growth cone motility.     

The axonal glycoprotein TAG-1 is an immunoglobulin superfamily member with neurite outgrowth-promoting activity

Pathfinding of axons in the developing nervous system is thought to be mediated by glycoproteins expressed on the surface of embryonic axons and growth cones. One molecule suggested to play a role in axonal growth is TAG-1, a 135 kd glycoprotein expressed transiently on the surface of subsets of neurons in the developing mammalian nervous system. We isolated a full-length cDNA clone encoding rat TAG-1. TAG-1 has six immunoglobulin-like domains and four fibronectin type III-like repeats and is structurally similar to other immunoglobulin-like proteins expressed on developing axons. Neurons maintained in vitro on a substrate of TAG-1 extend long neurites, suggesting that this protein plays a role in the initial growth and guidance of axons in vivo. TAG-1 is anchored to the neuronal membrane via a glycosyl phosphatidylinositol linkage and is also released from neurons, suggesting that TAG-1 also functions as a substrate adhesion molecule when released into the extracellular environment.     

Pathfinding by sensory axons in Drosophila: substrates and choice points in early lch5 axon outgrowth

We have examined the pattern of axon growth from the lateral chordotonal (lch5) neurons in the body wall of the Drosophila embryo and identified cellular substrates and choice points involved in early axon pathfinding by these sensory neurons. At the first choice point (TP1), the lch5 growth cones contact the most distal cells of the spiracular branch (SB) of the trachea. The SB provides a substrate along which the axons extend internally to the level of the intersegmental nerve (ISN). In the absence of the SB, the lch5 axons often stall near TP1 or follow aberrant routes towards the CNS. At the second choice point (TP2), the lch5 growth cones make their first contact with other axons and turn ventrally toward the CNS, fasciculating specifically with the motor axons of the ISN.     

The role of laminin and the laminin/fibronectin receptor complex in the outgrowth of retinal ganglion cell axons

Chick embryo retinal ganglion cell (RGC) axons grow to the optic tectum along a stereotyped route, as if responding to cues distributed along the pathway. We showed previously that, in culture, RGCs from embryonic Day 6 retina are responsive to the neurite-promoting effects of the extracellular matrix glycoprotein laminin and that this response is lost by RGCs at a later stage of development. Here we report that, before axon outgrowth is initiated in vivo, laminin, is expressed along the optic pathway at nonbasal lamina sites that are accessible to the growth cones of RGC axons. The distribution of laminin within the pathway is consistent with its localization at the end-feet of neuroepithelial cells that line the route, and it continues to be expressed at these marginal sites during the first week of embryonic development. At later stages, concomitant with the loss of response by RGCs in culture, laminin becomes restricted to basal laminae at the retinal inner limiting membrane and pial surface of the optic pathway. Neurofilament-positive RGC axons bind a monoclonal antibody, JG22, which recognizes the laminin/fibronectin receptor complex, and continue to do so throughout embryonic development. We show that, in vitro, the JG22 antigen expressed by RGCs appears to function as a laminin receptor, by demonstrating that JG22 antibody blocks neurite outgrowth on a substrate of laminin. These findings are consistent with the possibility that laminin defines a transient performed pathway specifically recognized by early RGC growth cones as they navigate toward their central target.     

Glia dictate pioneer axon trajectories in the Drosophila embryonic CNS

Whereas considerable progress has been made in understanding the molecular mechanisms of axon guidance across the midline, it is still unclear how the axonal trajectories of longitudinal pioneer neurons, which never cross the midline, are established. Here we show that longitudinal glia of the embryonic Drosophila CNS direct formation of pioneer axon pathways. By ablation and analysis of glial cells missing mutants, we demonstrate that glia are required for two kinds of processes. Firstly, glia are required for growth cone guidance, although this requirement is not absolute. We show that the route of extending growth cones is rich in neuronal cell bodies and glia, and also in long processes from both these cell types. Interactions between neurons, glia and their long processes orient extending growth cones. Secondly, glia direct the fasciculation and defasciculation of axons, which pattern the pioneer pathways. Together these events are essential for the selective fasciculation of follower axons along the longitudinal pathways.     

Glial interactions with neurons during Drosophila embryogenesis

A monoclonal antibody (Mab5B12) demonstrating specificity for glial cells within the central and peripheral nervous systems of Drosophila has been used in combination with neural-specific antibodies to study the early organization of the Drosophila embryo. The embryonic central nervous system of Drosophila contains cells within the ventral midline that are recognized by monoclonal antibody 5B12. These cells are not recognized by either a polyclonal antiserum to horse radish peroxidase, which recognizes several antigens on the surface of Drosophila neurons, or Mab22C10, which recognizes an antigen specific to the peripheral nervous system. Mab5B12-positive cells lie dorsal both to the developing anterior and posterior commissures in each thoracic and abdominal segment and to the supraoesophageal commissure. They ensheath these commissures in later stage embryos. Other Mab5B12-positive cells lie dorsolateral to the CNS and send processes laterally to the lateral sensilla during axonogenesis in the PNS. These cells surround the axons of the intersegmental and segmental nerves. Other cells that line the advancing ectoderm during dorsal closure and surround the anal pads also express the Mab5B12 antigen. Neuronal cell cultures derived from Drosophila gastrulae contain cells expressing the Mab5B12 antigen. These cells can be found separate or in close association with neuronal clusters and their axons.     

Human dorsal root ganglion neurons from embryonic donors extend axons into the host rat spinal cord along laminin-rich peripheral surroundings of the dorsal root transitional zone

Following dorsal root crush, the lesioned axons regenerate in the peripheral compartment of the dorsal root, but stop at the boundary between the peripheral and the central nervous system, the dorsal root transitional zone. We have previously shown that fibres from human fetal dorsal root ganglia grafted to adult rat hosts are able to grow into the spinal cord, but were not able to specify the route taken by the ingrowing fibres. In this study we have challenged the dorsal root transitional zone astrocyte boundary with human dorsal root ganglion transplants from 5–8-week-old embryos. By tracing immunolabelled human fibres in serial sections, we found that fibres consistently grow around the dorsal root transitional zone astrocytes in laminin-rich peripheral surroundings, and extend into the host rat spinal cord along blood vessels, either into deep or superficial laminae of the dorsal horn, or into the dorsal funiculus. Human fibres that did not have access to blood vessels grew on the spinal cord surface. These findings indicate, that in spite of a substantial growth capacity by axons from human embryonic dorsal root ganglion cells as well as their tolerance to non-permissive factors in the mature mammalian CNS, these axons are still sensitive to the repellent effects of astrocytes of the mature dorsal root transitional zone. Furthermore, this axonal ingrowth is consistently associated with laminin-expressing structures until the axons reach the host spinal cord.     

Spatial regulation of axonal glycoprotein expression on subsets of embryonic spinal neurons

The identification of surface proteins restricted to subsets of embryonic axons and growth cones may provide information on the mechanisms underlying axon fasciculation and pathway selection in the vertebrate nervous system. We describe here the characterization of a 135 kd cell surface glycoprotein, TAG-1, that is expressed transiently on subsets of embryonic spinal cord axons and growth cones. TAG-1 is immunochemically distinct from the cell adhesion molecules N-CAM and L1 (NILE) and is expressed on commissural and motor neurons over the period of initial axon extension. Moreover, TAG-1 and L1 appear to be segregated on different segments of the same embryonic spinal axons. These observations provide evidence that axonal guidance and pathway selection in vertebrates may be regulated in part by the transient and selective expression of distinct surface glycoproteins on subsets of developing neurons.     

Morphology of pioneer and follower growth cones in the developing cerebral cortex

In the developing nervous systems of both invertebrates and vertebrates, neurons must develop precise sets of axonal connections. One strategy used by both orders of animals is to generate a special class of neurons whose axons "pioneer" the first pathways between these cells and their targets. In the developing mammalian telencephalon, the subplate neurons (which are among the first neurons to be generated in development) extend axons to long-distance subcortical targets before the neurons of the deep cortical layers 5 and 6 have been generated. The axons of layer 5 and 6 neurons later follow a similar pathway to form permanent subcortical projections to the thalamus and tectum, and thereafter the vast majority of subplate neurons die. These results have generated the hypothesis that subplate axons may actually be required for the axons of layer 5 and 6 neurons to innervate their appropriate subcortical targets. The complexity of growth cones has previously been correlated with axonal decision making: differences in growth cone morphologies have been noted in comparisons of leading versus following axons (LoPresti, Macagno, and Levinthal, 1973; Nordlander, 1987; Yaginuma, Homma, Kunzi, and Oppenheim, 1991), and at choice points along axon pathways (Raper, Bastiani, and Goodman, 1983; Tosney and Landmesser, 1985; Caudy and Bentley, 1986a,b; Bovolenta and Mason, 1987; Holt, 1989; Bovolenta and Dodd, 1990; Yaginuma et al., 1991). Thus, as a first step toward addressing the question of whether the axons of deep-layer neurons simply follow subplate axons to their targets, we have studied the morphology of cortical growth cones at various points along the corticothalamic pathway and at different stages of development. We examined the brains of fetal ferrets and cats at ages ranging from embryonic days (E) 24 to E50, using the fluorescent lipophilic tracer 1,1-dioctadecyl-3,3,3',3'-tetramethyl indocarbocyanine perchlorate (DiI) to reveal the axons and growth cones of cortical neurons. Growth cones were drawn, and quantitative measurements of their complexity were made by counting filopodia and calculating their surface area. No morphological differences were found among growth cones at different points along the corticothalamic pathway at a given age. However, growth cones belonging to early-generated cells (likely to be subplate neurons) are significantly larger and more complex than are the growth cones of later-generated cortical neurons. This evidence is consistent with the suggestion that subplate growth cones actively pioneer the corticothalamic pathway, and that the axons of layer 5 and 6 neurons follow it.