Christopher S. Cohan

Calcium regulation of the neuronal growth cone

Abstract The growth cone behaviors that are involved in the generation of neuronal cytoarchitecture are apparently regulated in quite specific ways by Ca 2+ . Neurotransmitters and electrical activity, well known for their roles in information coding, have recently been shown to affect growth cone motility by mechanisms linked to Ca 2+ . Ca 2+ may therefore act as a common integrator of environmental cues that influence neurite outgrowth and synaptogenesis, and in this way may play a key role in the establishment and modulation of brain circuitry.

Focal loss of actin bundles causes microtubule redistribution and growth cone turning

Ît is commonly believed that growth cone turning during pathfinding is initiated by reorganization of actin filaments in response to guidance cues, which then affects microtubule structure to complete the turning process. However, a major unanswered question is how changes in actin cytoskeleton are induced by guidance cues and how these changes are then translated into microtubule rearrangement. Here, we report that local and specific disruption of actin bundles from the growth cone peripheral domain induced repulsive growth cone turning. Meanwhile, dynamic microtubules within the peripheral domain were oriented into areas where actin bundles remained and were lost from areas where actin bundles disappeared. This resulted in directional microtubule extension leading to axon bending and growth cone turning. In addition, this local actin bundle loss coincided with localized growth cone collapse, as well as asymmetrical lamellipodial protrusion. Our results provide direct evidence, for the first time, that regional actin bundle reorganization can steer the growth cone by coordinating actin reorganization with microtubule dynamics. This suggests that actin bundles can be potential targets of signaling pathways downstream of guidance cues, providing a mechanism for coupling changes in leading edge actin with microtubules at the central domain during turning.

Sealing cultured invertebrate neurons to embedded dish electrodes facilitates long-term stimulation and recording

Recently it has become possible to form small networks of synaptically connected identified invertebrate neurons in culture. Using conventional saline-filled glass electrodes, it is difficult to simultaneously stimulate and record from more than 2 or 3 cultured neurons and to perform experiments lasting longer than several hours. We demonstrate that it is possible to overcome these limitations by using planar arrays of electrodes embedded in the bottom of a culture dish. The arrays employ conductive leads and insulation that are transparent, making the dishes compatible with voltage-sensitive dyes and inverted microscopy. Identified neurons from leech Hirudo medicinalis, slug Aplysia californica, and snail Helisoma trivolvis, have been grown on these arrays. Due to their large size (soma diameter 40-200 microns) these neurons form seals over the dish electrodes. Individual electrodes can then be used to stimulate and to record action potentials in the associated neuron. With sealing, action potentials have been recorded simultaneously from many neurons for up to two weeks, with signal-to-noise ratios as large as 500:1. We developed and tested a simple model that describes the voltage waveforms measured with array electrodes. Potentials measured from electrodes under cell bodies were primarily derivatives of the intracellular potential, while those measured from electrodes under axon stumps were primarily proportional to local inward Na+ currents. While it is relatively easy to record action potentials, it is difficult to record postsynaptic potentials because of their small size and slow rate of rise.

Interactive Effects of Serotonin and Acetylcholine on Neurite Elongation

Serotonin (5-HT) inhibits elongation of neurites of specific identified neurons. Here we report a novel, growth-enabling action of another neurotransmitter, acetylcholine (ACh). When applied simultaneously with serotonin, ACh prevents the inhibition of Helisoma neuron B19 neurite elongation that would occur in response to application of 5-HT alone. We also report that ACh prevents the rise in growth cone Ca2+ that would occur in response to application of 5-HT alone and that ACh blocks the electrical excitatory effect of 5-HT on neuron B19. These results support the hypothesis that growth cone motility and neurite elongation can be regulated by voltage-gated Ca2+ fluxes and suggest that the dynamics of neurite morphology may be complexly regulated by an array of neurotransmitters, as is functional electrical activity.

Role of the actin bundling protein fascin in growth cone morphogenesis: localization in filopodia and lamellipodia.

Growth cones at the distal tips of growing nerve axons contain bundles of actin filaments distributed throughout the lamellipodium and that project into filopodia. The regulation of actin bundling by specific actin binding proteins is likely to play an important role in many growth cone behaviors. Although the actin binding protein, fascin, has been localized in growth cones, little information is available on its functional significance. We used the large growth cones of the snail Helisoma to determine whether fascin was involved in temporal changes in actin filaments during growth cone morphogenesis. Fascin localized to radially oriented actin bundles in lamellipodia (ribs) and filopodia. Using a fascin antibody and a GFP fascin construct, we found that fascin incorporated into actin bundles from the beginning of growth cone formation at the cut end of axons. Fascin associated with most of the actin bundle except the proximal 6--12% adjacent to the central domain, which is the region associated with actin disassembly. Later, during growth cone morphogenesis when actin ribs shortened, the proximal fascin-free zone of bundles increased, but fascin was retained in the distal, filopodial portion of bundles. Treatment with tumor promoter 12-O-tetradecanoylphorbol-13-acetate (TPA), which phosphorylates fascin and decreases its affinity for actin, resulted in loss of all actin bundles from growth cones. Our findings suggest that fascin may be particularly important for the linear structure and dynamics of filopodia and for lamellipodial rib dynamics by regulating filament organization in bundles.

A high-speed multispectral spinning-disk confocal microscope system for fluorescent speckle microscopy of living cells

Fluorescent speckle microscopy (FSM) uses a small fraction of fluorescently labeled subunits to give macromolecular assemblies such as the cytoskeleton fluorescence image properties that allow quantitative analysis of movement and subunit turnover. We describe a multispectral microscope system to analyze the dynamics of multiple cellular structures labeled with spectrally distinct fluorophores relative to one another over time in living cells. This required a high-resolution, highly sensitive, low-noise, and stable imaging system to visualize the small number of fluorophores making up each fluorescent speckle, a means by which to switch between excitation wavelengths rapidly, and a computer-based system to integrate image acquisition and illumination functions and to allow a convenient interface for viewing multispectral time-lapse data. To reduce out-of-focus fluorescence that degrades speckle contrast, we incorporated the optical sectioning capabilities of a dual-spinning-disk confocal scanner. The real-time, full-field scanning allows the use of a low-noise, fast, high-dynamic-range, and quantum-efficient cooled charge-coupled device (CCD) as a detector as opposed to the more noisy photomultiplier tubes used in laser-scanning confocal systems. For illumination, our system uses a 2.5-W Kr/Ar laser with 100-300mW of power at several convenient wavelengths for excitation of few fluorophores in dim FSM specimens and a four-channel polychromatic acousto-optical modulator fiberoptically coupled to the confocal to allow switching between illumination wavelengths and intensity control in a few microseconds. We present recent applications of this system for imaging the cytoskeleton in migrating tissue cells and neurons.

Actin dynamics and organization during growth cone morphogenesis in Helisoma neurons

Growth cone formation at the terminal region of severed axons is a fundamental step in neuronal regeneration. To understand the cytoskeletal events underlying this process, we have followed actin organization and dynamics as the severed, axonal stumps of Helisoma neurons transformed into mature growth cones. We identified three stages in growth cone morphogenesis: (1) formation, (2) expansion, and (3) maturation. The formation stage involved cytochalasin B-insensitive terminal swelling formation, followed by cytochalasin B-inhibited filopodial and lamellipodial formation. Time-lapse images of neurons injected with labeled actin showed actin ribs in nascent growth cones formed both by incorporation of filopodial actin bundles and de novo assembly at the leading edge. Phallacidin-stained growth cones revealed F-actin to be organized into bundles (ribs) and a meshwork throughout morphogenesis. Actin ribs represented the dominant F-actin population during the expansion stage and the early phase of maturation, whereas a meshwork organization dominated the late phase of maturation. During the expansion stage, growth cones exhibited a rapid retrograde flow (4.8 microns/min), as assessed with flow-coupled latex beads, and comparatively slow lamellipodial protrusion (0.3 micron/min). During the maturation stage, no net lamellipodial advancement occurred; however, the rate of retrograde flow was significantly faster in the early phase (5.0 microns/min) than the late phase (2.3 microns/min). This decrease in retrograde flow corresponded with a change in actin organization. Lateral movements of actin ribs (2.1 microns/min) also occurred throughout growth cone morphogenesis, but were most prominent during the expansion stage. These experiments provide evidence for de novo actin assembly during growth cone formation and demonstrate that temporal changes in actin organization and dynamics accompany growth cone morphogenesis.

Developmental regulation of a neurite-promoting factor influencing statoacoustic neurons.

The present study investigated a target-derived, neurite-promoting factor (NPF) released by the developing chick otocyst and its effects on statoacoustic ganglia (SAG). SAG explants cultured in the absence of otocysts produced little neurite outgrowth at all stages of development examined (E4-E13). However, extensive neurite outgrowth was seen when E4-E6 SAG were cultured in the presence of otocysts of the same age. The amount of neurite outgrowth observed in cocultures steadily decreased at later developmental stages. E7-E9 cocultures produced less outgrowth and E10-E13 cocultures produced the least outgrowth compared to E4-E6 cocultures. Additionally, otocysts from older stages were unable to promote outgrowth of E4 SAG. Thus, the level of the factor released by the otocysts declined during development. In contrast, neurite outgrowth was promoted when E10-E15 SAG were cocultured in the presence of younger stage otocysts. Our data indicate that the release of NPF from chick otocysts decreased from E6 to E13, although the ability of SAG neurons to respond to the NPF was maintained throughout development.

Ca2+ dynamics in neuronal growth cones: regulation and changing patterns of Ca2+ entry.

Digital ratio imaging of Fura-2 fluorescence was used to determine spatially resolved dynamics of Ca2+ changes in neuronal growth cones from the molluscs, Helisoma and Aplysia. Time resolution was approximately 1 s and spatial resolution a few mm depending upon the thickness of the cell region examined. Isolated growth cones of Helisoma were shown to recover from large Ca2+ loads over a time course of minutes, therefore demonstrating Ca2+ regulation mechanisms not dependent on the rest of the cell. Ca2+ changes monitored during action potential discharge showed sharply defined spatial gradients within the growth cones, probably arising from clustering of voltage-gated Ca-channels in the surface membrane. The regions of peak concentration change appeared to shift from central regions to the growth cone periphery as the growth cones matured. There was a marked difference in soma Ca2+ changes produced by action potentials depending on whether or not the soma had sprouted neurites. Neurite-free somata showed large Ca2+ changes, whereas in somata that had recently sprouted neurites there were almost no changes for similar electrical stimulation. Measurements on growth cones of N1E115 neuroblastoma cells showed static distributions of Ca2+ similar to those in the molluscan neurons.

F-actin at Newly Invaginated Membrane in Neurons: Implications for Surface Area Regulation

Abstract. Neuronal shape and volume changes require accompanying cell surface adjustments. In response to osmotic perturbations, neurons show evidence of surface area regulation; shrinking neurons invaginate membrane at the substratum, pinch off vacuoles, and lower their membrane capacitance. F-actin is implicated in reprocessing newly invaginated membrane because cytochalasin causes the transient shrinking-induced invaginations, vacuole-like dilations (VLDs), to persist indefinitely instead of undergoing recovery. To help determine if cortical F-actin indeed contributes to cell surface area regulation, we test, here, the following hypothesis: invaginating VLD membrane rapidly establishes an association with F-actin and this association contributes to VLD recovery. Cultured molluscan (Lymnaea) neurons, whose large size facilitates three-dimensional imaging, were used. In fixed neurons, fluorescent F-actin stains were imaged. In live neurons, VLD membrane was monitored by brightfield microscopies and actin was monitored via a fluorescent tag. VLD formation (unlike VLD recovery) is cytochalasin insensitive and consistent with this, VLDs formed readily in cytochalasin-treated neurons but showed no association with F-actin. Normally, however (i.e., no cytochalasin), VLDs were foci for rapid reorganization of F-actin. At earliest detection (1–2 min), nascent VLDs were entirely coated with F-actin and by 5 min, VLD mouths (i.e., at the substratum) had become annuli of F-actin-rich motile leading edge. Time lapse images from live neurons showed these rings to be motile filopodia and lamellipodia. The retrieval of VLD membrane (vacuolization) occurred via actin-associated constriction of VLD mouths. The interplay of surface membrane and cortical cytoskeleton in osmotically perturbed neurons suggests that cell surface area and volume adjustments are coordinated in part via mechanosensitive F-actin dynamics.