Daniel J. Goldberg

Cytoplasmic dynein and LIS1 are required for microtubule advance during growth cone remodeling and fast axonal outgrowth

Recent evidence has implicated dynein and its regulatory factors dynactin and LIS1 in neuronal and non-neuronal cell migration. In the current study we sought to test whether effects on neuronal cell motility might reflect, in part, a role for these proteins in the growth cone. In chick sensory neurons subjected to acute laminin treatment dynein, dynactin, and LIS1 were mobilized strikingly and rapidly to the leading edge of the growth cone, where they were seen to be associated with microtubules converging into the laminin-induced axonal outgrowths. To interfere acutely with LIS1 and dynein function and to minimize secondary phenotypic effects, we injected antibodies to these proteins just before axon initiation. Antibody to both proteins produced an almost complete block of laminin-induced growth cone remodeling and the underlying reorganization of microtubules. Penetration of microtubules into the peripheral zone of differentiating axonal growth cones was decreased dramatically by antibody injection, as judged by live analysis of enhanced green fluorescent protein-tubulin and the microtubule tip-associated EB3 (end-binding protein 3). Dynein and LIS1 inhibition had no detectable effect on microtubule assembly but reduced the ability of microtubules to resist retrograde actin flow. In hippocampal neurons dynein, dynactin, and LIS1 were enriched in axonal growth cones at stage 3, and both growth cone organization and axon elongation were altered by LIS1 RNA interference. Together, our data indicate that dynein and LIS1 play a surprisingly prominent role in microtubule advance during growth cone remodeling associated with axonogenesis. These data may explain, in part, the role of these proteins in brain developmental disease and support an important role in diverse aspects of neuronal differentiation and nervous system development.

Rapid effects of laminin on the growth cone

To gain insight into how laminin promotes neurite growth, high resolution video microscopy was used to determine the rapid effects of laminin on growth cone structure. Sympathetic growth cones in serum-free medium on polylysine substrate displayed extensive motility and protrusive activity and often had large lamellipodia. However, their neurites grew slowly because membranous organelles from the central region advanced into the lamellipodium only slowly. Acute addition of laminin accelerated growth severalfold and had visible effects on the growth cone within minutes. Laminin dramatically accelerated the advance of membranous organelles, which, with microtubules, rapidly filled the lamellipodium. Retraction of individual protrusions (filopodia and veils) was rapidly reduced. These effects of laminin are important in accelerating growth and suggest a mechanism for pathway selection by growing neurites.

Crosstalk between p38, Hsp25 and Akt in spinal motor neurons after sciatic nerve injury

The p38 stress-activated protein kinase pathway is involved in regulation of phosphorylation of Hsp25, which in turn regulates actin filament dynamic in non-neuronal cells. We report that p38, Hsp25 and Akt signaling pathways were specifically activated in spinal motor neurons after sciatic nerve axotomy. The activation of the p38 kinase was required for induction of Hsp25 expression. Furthermore, Hsp25 formed a complex with Akt, a member of PI-3 kinase pathway that prevents neuronal cell death. Together, our observations implicate Hsp25 as a central player in a complex system of signaling that may both promote regeneration of nerve fibers and prevent neuronal cell death in the injured spinal cord.

Looking into growth cones

The growth cone is a crucial structure in effecting neurite elongation and guiding the neurite onto correct pathways by responding to environmental cues. Recently developed techniques in light and electron microscopy have greatly improved our understanding of the dynamic organization of membrane and cytoskeleton within the growth cone. The growth cone can now be directly observed to undergo a sequence of developmental changes to produce the neurite. The importance, in elongation and steering, of pulling of growth cone protrusions against adhesive contacts on the substrate is re-evaluated in the light of these findings.

Micropruning: the mechanism of turning of Aplysia growth cones at substrate borders in vitro

Growth cones of Aplysia californica neurons were observed with video- enhanced contrast-differential interference contrast (VEC-DIC) microscopy as they turned at a border between poly-L-lysine-treated and untreated glass. Growth cones that turned generally developed 2 distinct active areas of filopodial and veil formation, much in the way of growth cones undergoing branching. Both active areas advanced, but turning of the neurite occurred through the selective resorption of the incipient branches developing on the untreated substrate. Thus, micropruning of developing regions of the growth cone, rather than the asymmetric extension of filopodia or veils, was primarily responsible for directing neurite growth. We present the hypothesis that abrupt turns by growing neurites are mediated by 2 sets of signals, one causing growth cone splitting, and a second set regulating the survival of the separate branches.

Differential growth of the branches of a regenerating bifurcate axon is associated with differential axonal transport of organelles.

Axonal trees display differential growth during development or regeneration; that is, some branches stop growing and often retract while other branches continue to grow and form stable synaptic connections. In this study, an in vitro model of differential growth is examined to identify the intracellular events responsible for this phenomenon. When the giant cerebral neuron of Aplysia californica is placed in culture, vigorous growth occurs from the ends of both branches of its bifurcate axon. If an appropriate target neuron is placed next to one branch, growth from that branch is unabated while growth from the other branch is suppressed. The bidirectional fast transport of membranous organelles was examined in the two branches by the use of high-resolution video microscopy. Transport was similar in the branches in the absence of a target cell but was much greater in the growing than in the nongrowing branch when a target was present. Electron microscopic examination of fixed specimens confirmed these findings. Differential growth may be initiated or sustained by a diversion from certain branches of materials used in growth which are supplied by fast axonal transport.

Microtubule and Rac 1-dependent F-actin in growth cones

Extracellular cues control the rate and direction of growth of neuronal processes in large part by regulating the cytoskeleton of the growth cone. The actin filament network of the peripheral region is thought to be the primary target for these cues, with consequences for the advance and organization of microtubules. Binding of laminin to integrin receptors is a cue that accelerates the growth of processes from many types of neurons. It was applied acutely to sympathetic neurons in culture to study its effects on the cytoskeleton of the growth cone. Microtubules advance to the edge of the growth cone and bundle in response to laminin, and it was found that small veils of membrane appear near the ends of some of those microtubules. To examine more clearly the relationship between the microtubules and the appearance of actin-rich structures at the periphery, a low dose of cytochalasin D was used to deplete the peripheral region of the growth cone of pre-existing F-actin. The subsequent addition of laminin resulted in the bundling of ends of dynamic (tyrosinated) microtubules at the distal edge of the growth cone, most of which were associated with foci of F-actin. Observations of labeled actin within living growth cones confirmed that these foci formed in response to laminin. Suppression of microtubule dynamics with drugs eliminated the actin foci; washout of drug restored them. Rac 1 did not co-concentrate with F-actin in the peripheral region of the growth cone in the absence of laminin, but did co-concentrate with the foci of F-actin that formed in response to laminin. Inhibition of Rac 1 functioning prevented the formation of the foci and also inhibited laminin-induced neurite growth with or without cytochalasin. These results indicate that extracellular cues can affect actin in the growth cone via microtubules, as well as affect microtubules via actin. They also point to the mediation of microtubule-dependent accumulation of F-actin at the front of the growth cone as a role of Rac 1 in neurite growth.

Recruitment of the Arp2/3 complex and mena for the stimulation of actin polymerization in growth cones by nerve growth factor.

The growth of axons and dendrites during development and regeneration is regulated by cues in the environment. Many of these cues regulate the actin cytoskeleton of the protrusive structures (like filopodia) of the growth cone that are essential for detecting and responding to cues. Nerve growth factor, which promotes the formation of protrusive structures, stimulated actin polymerization in rat sympathetic growth cones, resulting within 1–2 min in accumulations of F‐actin at the distal edge and in splotches of F‐actin farther back. We examined the potential involvement of a protein machinery important in at least certain types of actin polymerization in non‐neuronal cells. Members of the Arp2/3 complex, p34‐Arc and p21‐Arc, heavily concentrated in the early accumulations of F‐actin, as did one member of the Ena/VASP family (Mena) but not another (VASP). Retention of Arc proteins at preferred sites of actin polymerization did not require polymerization itself. Growth cones of differentiated PC12 cells were similar to sympathetic growth cones in their response to NGF. Introduction into these cells of a peptide that should block the binding of Ena/VASP family proteins to the protein complex at sites of actin polymerization reduced the formation of splotches and filopodia in response to NGF. These results point to the early involvement of the Arp2/3 complex and the Ena/VASP family in growth factor‐stimulated actin polymerization that gives rise to protrusive structures at the growth cone. J. Neurosci. Res. 4:458–467, 2000

Braking news: calcium in the growth cone.

The critical unknown was whether signaling through Ca1 in the growth cone is physiologically relevant. It now seems likely that it is. One line of evidence is its Daniel J. Goldberg* and Peter W. Grabham Department of Pharmacology and Center for Neurobiology and Behavior apparent involvement in transducing responses in the Columbia University growth cone to certain cues that are likely to function New York, New York 10032 in vivo. NI-35 is a growth-inhibitory protein expressed on oligodendrocytes in the CNS (see Bandtlow et al., 1993, for references). Regeneration of injured axons Growing axons are guided by cues in the environment: within the spinal cord is promoted by antibodies that local or released from a distance, positive and negative neutralize NI-35, attesting to its importance in vivo. NI(Goodman, 1996). The motile growth cone at the tip of 35 induces growth cones in culture to collapse. The the axon senses and interprets these cues, changing its collapse is associated with a large rise in [Ca]i, at least behavior to alter the direction or speed of growth. Many partially due to release from the smooth endoplasmic cues have been identified and their in vivo importance reticulum (SER) (Bandtlow et al., 1993). Blockage of Ca1 has been verified, but how their binding is transduced release prevents collapse. into changes in growth cone behavior remains poorly Certain positive cues that are likely to function in vivo understood. Given the plethora of cues, it would simplify may also work through Ca1. NCAM and L1 are growththings if common elements of transduction pathways promoting members of the immunoglobulin superfamily emerged: for example, if stopping of a growth cone at (Walsh and Doherty, 1997). Disruption of the functioning a choice point were always (or, at least, often) caused of either in vivo leads to what appear to be abnormalities by elevation of messenger X within the growth cone, in axonal growth (Cohen et al., 1998; Seki and Rutiswhatever the external cue. Recent work points to such hauser, 1998). The promotion of growth by either in vitro key transduction molecules. Moreover, a new paper in depends on influx of Ca1 through channels in the Nature literally and figuratively illuminates the inner plasma membrane, though it is merely assumed these workings of the growth cone by fluorescently imaging are in the growth cone (Walsh and Doherty, 1997). LamiCa1 in neurons growing within the developing spinal nin is a prominent constituent of the extracellular matrix cord, providing evidence for its importance in vivo in and is expressed by Schwann cells in peripheral nerve, regulating axonal elongation (Gomez and Spitzer, 1999). where it probably promotes growth. It elicits rises in Considering that the response of growth cones to cues [Ca]i in growth cones in culture that may be important is, for technical reasons, typically studied in the reducin affecting both the rate and direction of growth (Kuhn tionist world of the culture dish, this is a noteworthy et al., 1998; but see below). accomplishment. The aforementioned environmental cues all act as Calcium Is a Physiologically Important Transduction substrate-bound molecules. Molecules released from Molecule in the Growth Cone afar can also influence growth, either attracting or repelIt has long been known that the growth cone is sensitive ling neurites. Here, Ca1 may also be involved. Netrin-1 to its [Ca]i. Large increases caused by neurotransmitis produced by the floor plate of the developing spinal ters or depolarization cause growth cones in culture to cord, from where it emanates to attract certain axons collapse, stopping neuritic elongation (Kater and Mills, to, and repel others from, the midline (see Ming et al., 1991). Decreases caused by removing Ca1 from the 1997, for references). This Janus-like behavior can be bathing medium can have similar effects. In some cases, displayed in culture in a particularly intriguing way. A growth cone activity and neuritic elongation can be prospinal neurite that turns and grows toward a micropimoted by elevations of [Ca]i over resting levels; focal pette emitting netrin-1 turns away from that micropipette changes within the growth cone can produce focal prowhen the activity of cAMP-dependent protein kinase in trusive activity appropriate for turning, for example (Davthe neurite is pharmacologically suppressed (Ming et enport and Kater, 1992). These seemingly discordant al., 1997). Neither behavior is seen when [Ca]o is drastiresults were accommodated by Kater’s set-point model, cally reduced, though straight-ahead growth proceeds in which substantial deviations in either direction from apace. Similar results are seen with brain-derived neuroan optimal [Ca]i inhibit motile activity of the growth trophic factor (BDNF) and acetylcholine. A complemencone and slow growth (Kater and Mills, 1991). This is tary situation is seen with myelin-associated glycoproso reasonable as to be unsurprising; very high or low tein (MAG), another inhibitory protein expressed by levels of [Ca]i are cytotoxic, presumably one reason oligodendrocytes (see Song et al., 1998, for references). that [Ca]i is normally tightly buffered. However, there Neurites that are repelled by a distant source of MAG were indications that the deviations from normal did are instead attracted when cAMP is artificially elevated not have to be great to inhibit growth (Lankford and (Song et al., 1998). Both the repulsion and the attraction Letourneau, 1991). Microfilaments, so important to the disappear in low [Ca1] medium. motile activity of the growth cone, are particularly sensiAlso pointing to a role for Ca1 in regulating neuritic tive to [Ca]i. growth was the discovery that certain growth cones advancing in culture display periodic brief rises in [Ca]i (termed waves) that seem to affect the speed of neuritic * To whom correspondence should be addressed (e-mail: dlg5@

Bundling of Microtubules in the Growth Cone Induced by Laminin

Axons growing in the developing nervous system are guided by cues in the environment which act at the growth cone. So far, the initial cytoskeletal target of these cues has been found to be the network of actin filaments in the peripheral region of the growth cone. Laminins are constituents of the extracellular matrix which promote axonal growth. They exert effects on the actin network. Here, laminin 1 is shown to affect microtubules as well. Acute addition of laminin 1 to rat sympathetic neurons quickly caused the advance of microtubules and their bundling within the initial widely spread growth cone and then the outgrowth of thin, rapidly growing nascent axons. The bundling was pharmacologically separable from the advance of microtubules caused by laminin, as the former but not the latter was blocked by lithium. The bundling did not depend on the peripheral network of actin filaments, as it was unimpaired by the removal of this network with cytochalasin D. Thus, microtubules seem to be a direct cytoskeletal target for laminin 1 in the growth cone, with important consequences for axonal outgrowth.