Thy-1 antibody-triggered neurite outgrowth requires an influx of calcium into neurons via N- and L-type calcium channels.

We present evidence that the neurite out-growth stimulated by the binding of Thy-1 antibodies to PC12 cells is mediated by calcium influx through both N- and L-type calcium channels. PC12 cells cultured on a noncellular substratum in the presence of NGF, or on a cellular substratum in the absence of NGF, responded to soluble Thy-1 antibody by extending longer neurites. The response required bivalent antibody and could be blocked by removing Thy-1 from the surface of PC12 cells with phosphatidylinositol specific phospholipase C. The response could also be blocked by reducing extracellular calcium to 0.25 mM, or by antagonists of L- and N-type calcium channels. Additionally, the response could be fully inhibited by preloading PC12 cells with BAPTA/AM which buffers changes in intracellular calcium. A heterotrimeric G-protein is also implicated in the pathway as the response could be fully inhibited by pertussis toxin. These data suggest that antibody-induced clustering of Thy-1 stimulates neurite outgrowth by activating a second messenger pathway that has previously been shown to underlie cell adhesion molecule (NCAM, N-cadherin, and L1), but not integrin or NGF-dependent neurite outgrowth.

PDGF and intracellular signaling in the timing of oligodendrocyte differentiation.

In the rat optic nerve, bipotential O-2A progenitor cells give rise to oligodendrocytes and type 2 astrocytes on a precise schedule. Previous studies suggest that PDGF plays an important part in timing oligodendrocyte development by stimulating O-2A progenitor cells to proliferate until they become mitotically unresponsive to PDGF, stop dividing, and differentiate automatically into oligodendrocytes. Since the loss of mitotic responsiveness to PDGF has been shown not to be due to a loss of PDGF receptors, we have now examined the possibility that the unresponsiveness results from an uncoupling of these receptors from early intracellular signaling pathways. We show that (a) although PDGF does not stimulate newly formed oligodendrocytes to synthesize DNA, it induces an increase in cytosolic Ca2+ in these cells; (b) a combination of a Ca2+ ionophore plus a phorbol ester mimics the effect of PDGF, both in stimulating O-2A progenitor cell division and in reconstituting the normal timing of oligodendrocyte differentiation in culture; and (c) the same combination of drugs does not stimulate newly formed oligodendrocytes to proliferate, even in the presence of PDGF or dibutyryl cAMP. The most parsimonious explanation for these results is that O-2A progenitor cells become mitotically unresponsive to PDGF because the intracellular signaling pathways from the PDGF receptor to the nucleus are blocked downstream from the receptor and some of the early events that are triggered by receptor activation.

Two components of voltage-dependent calcium influx in mouse neuroblastoma cells. Measurement with arsenazo III.

N1E-115 mouse neuroblastoma cells were injected with the calcium indicator dye arsenazo III. Optical absorbance changes during voltage-clamp depolarization were used to examine the properties of the two calcium currents present in these cells. The rapidly inactivating calcium current (Moolenar and Spector, 1979b, Journal of Physiology, 292:307-323) inactivates by a voltage-dependent mechanism. The slowly inactivating calcium current is dominant in raising intracellular calcium during depolarizations to greater than -20 mV. Lowering the extracellular calcium concentration affects the two calcium currents unequally, with the slowly inactivating current being reduced more. Intracellular calcium falls very slowly (tau greater than 1 min) after a depolarization. The rapidly inactivating calcium current is responsible for a calcium action potential under physiological conditions. In contrast, it is unlikely that the slowly inactivating calcium current has an important electrical role. Rather, its function may be to add a further increment of calcium influx over and above the calcium influx through the rapidly inactivating calcium channels.

Calcium-dependent potassium current in barnacle photoreceptor.

When barnacle lateral eye photoreceptors are depolarized to membrane potentials of 0 to +50 mV in the dark, the plot of outward current through the cell membrane against time has two distinct maxima. The first maximum occurs 5-10 ms after the depolarization began. The current then decays to a minimum at approximately 500 ms after the onset of depolarization, and then increases to a second maximum 4-6 s after the depolarization began. If depolarization is maintained, the current again decays to reach a steady value approximately 1 min after depolarization began. The increase in current to the maximum at 4-6s from the minimum at approximately 500 ms is termed the "late current." It is maximum for depolarizations to around +25 mV and is reduced in amplitude at more positive potentials. It is not observed when the membrane is depolarized to potentials more positive than +60 mV. The late current is inhibited by external cobaltous ion and external tetraethylammonium ion, and shows a requirement for external calcium ion. When the calcium-sequestering agent EGTA is injected, the late current is abolished. Illumination of a cell under voltage clamp reduces the amplitude of the late current recorded subsequently in the dark. On the basis of the voltage dependence and pharmacology of the late current, it is proposed that the current is a calcium-dependent potassium current.

Nuclear calmodulin responds rapidly to calcium influx at the plasmalemma.

We have studied the rate and extent of calcium binding to calmodulin in neuronal cytosol and nucleus during brief calcium influx across the plasmalemma. Rat sensory neurones were whole-cell patch clamped using a pipette containing a fluorescent analogue of calmodulin that reports when it has bound calcium. Cytosolic and nuclear signals were separated using a confocal microscope. Plasmalemmal calcium influx during a one second depolarization that activated L type calcium channels caused large fractions of calmodulin in both the cytosol and nucleus to bind calcium. Thus, contrary to previous predictions, nuclear processes that require the calcium:calmodulin complex will be activated readily by even brief cell stimulation.

Spatial gradients of cytosolic calcium concentration in neurones during paradoxical activation by calcium.

4-Br-A23187 caused a calcium influx into chick sensory neurones and raised cytosolic calcium from a rest level of 97 +/- 7 nM to a peak of 296 +/- 30 nM. Despite the continued presence of ionophore, however, cytosolic calcium concentrations then fell. After 30 min in ionophore, cytosolic calcium concentration had returned to 105 +/- 5 nM, not significantly different from the value before ionophore addition. The permeability of the plasmalemma to divalent cations, as estimated by the manganese quench technique, was no lower at 30 min than at the peak of the cytosolic calcium transient. Thus the fall of calcium from its peak was not due to a slowing of calcium influx, but was due to an upregulation of mechanisms that remove calcium from the cytosol- an upregulation that persists even though cytosolic calcium has apparently returned to pre-stimulus levels. We used a novel fixed slit confocal microscope to examine the calcium concentration profile close to the plasmalemma. We found that after 25-30 min ionophore treatment, calcium concentration was elevated only in the cytoplasm within 1 micron of the plasmalemma. A maintained, elevated calcium under the plasmalemma can help explain the phenomenon of paradoxical activation seen in this and other cell types.

Characterization of Ca(2+)-activated 86Rb+ fluxes in rat C6 glioma cells: a system for identifying novel IKCa-channel toxins.

1. The pharmacological characteristics of a putative Ca2+ activated K+ channel (IKCa channel) in rat glioma C6 cells were studied in the presence of the Ca2+ ionophore, ionomycin and various K+ channel blockers, 86Rb+ being used as a radioisotopic tracer for K+. 2. The resting 86Rb+ influx into C6 cells was 318 +/- 20 pmol s-1. The threshold for ionomycin activation of 86Rb+ influx was approx. 100 nM. At ionomycin concentrations above the activation threshold, the initial rate of 86Rb+ influx was proportional to ionophore concentration. Ionomycin-activated 86Rb+ flux was saturable (EC50 = 0.62 +/- 0.03 microM) and was not inhibited by ouabain. 3. Intracellular Ca2+ increased within 30 s from a basal level of 42 +/- 2 nM to 233 +/- 17 nM, after addition of 2 microM ionomycin. During this period, intracellular pH fell from 7.03 +/- 0.04 to 6.87 +/- 0.03 and the cell hyperpolarized from -34 +/- 10 mV to -76 +/- 2 mV. 4. Single channel conductance measurements on inside-out patches in physiological K+ solutions identified a 14 +/- 3 pS CA(2+)-activated K+ current between -25 mV and +50 mV. In symmetrical (100 mM) K+, the single channel conductance was 26 pS. 5. Externally applied quinine (IC50 = 0.12 +/- 0.34 mM) and tetraethylammonium chloride (IC50 = 10 +/- 1.9 mM) inhibited 86Rb+ influx into C6 cells in a concentration-dependent manner. Charybdotoxin (IC50 = 0.5 +/- 0.02 nM) and iberiotoxin (IC50 = 800 +/- 150 nM), as well as the crude venoms from the scorpions Leiurus quinquestriatus and Mesobuthus tamulus, also inhibited 86Rb+ influx. In contrast, apamin and toxin I had no inhibitory effects on 86Rb+ flux. A screen of fractions from cation exchange h.p.l.c. of Mesob. tamulus venom revealed the presence of at least four charybdotoxin-like peptides. One of these was iberiotoxin; the other three are novel toxins. 6. The ionomycin-activated 86Rb+ influx into rat C6 glioma cells has proved to be a valuable pharmacological assay for the screening of toxins and crude venoms which modify intermediate conductance, Ca2+ activated K+ channel activity.

A nuclear location for Ca2+-activated adenylyl cyclases I and III in neurones.

Calcium/calmodulin-sensitive adenylyl cyclases are increasingly recognised as important factors in memory formation and synaptic plasticity. We have examined the distributions of adenylyl cyclases types I, III, and VIII within rat primary sensory neurons. Immunofluorescence revealed distinct staining for adenylyl cyclases type I and III, but not adenylyl cyclase type VIII, within the cell nucleus. Western blots suggest that a processed form of adenylyl cyclase type III may be found within primary neurons and PC12 cells as a 70-kDa protein. We propose that the observed nuclear adenylyl cyclases are soluble forms of the cyclases.

Biphasic effect of calcium on neurite outgrowth in neuroblastoma and cerebellar granule cells.

We have examined the effect of cytosolic calcium concentration ([Ca2+]i) on neurite outgrowth in two neuronal cells, cerebellar granule cells and N1E-115 neuroblastoma cells. The set point [Ca2+]i in unstimulated cells bathed in normal extracellular medium was 37 nM and 108 nM, respectively. When we altered extracellular calcium concentration to cause small excursions of [Ca2+]i either above or below the set point, neurite outgrowth from granule cells declined. Thus granule cells show the bell-shaped dependence of neurite outgrowth on [Ca2+]i characteristic of sensory and other neurones [Dev. Brain Res., 70 (1992) 287-290; The Axon, Oxford University Press, New York, 1994]. In contrast, neurite outgrowth from N1E-115 cells increased monotonically as [Ca2+]i was reduced. This result, which is consistent with results obtained by studying individual growth cones [J. Neurosci., 9 (1989) 4007-4020], implies that these transformed cells are aberrant in having no bell-shaped dependence of neurite outgrowth on [Ca2+]i. In both cell types an increase of [Ca2+]i above the set point reduced neurite outgrowth. However, this decline did not persist as [Ca2+]i was set to increasingly high levels by increasing extracellular calcium. Rather, in both cell types, an increase of extracellular calcium above 6.9 mM produced a second, increasing phase of neurite outgrowth.

Buffering intracellular calcium disrupts motoneuron development in intact zebrafish embryos.

Numerous studies, performed mainly on dissociated cells, have shown that calcium signals have a role during different stages of neuronal development. However, the actions of calcium during neuronal development in vivo remain to be established. The present study has investigated the role of intracellular calcium signals during development of motoneurons in the spinal cord of intact zebrafish embryos. Loading blastomeres of early embryos with either the calcium buffer BAPTA or the calcium reporter dye Calcium Green, was shown to disrupt motoneuron development in the spinal cord of embryos at 24 h postfertilisation. Loading the calcium buffer BAPTA, at an intracellular concentration of 1 mM, into the blastomeres of early embryos did not alter the resting levels of intracellular calcium, but significantly dampened transient rises in intracellular calcium in the cells of later stage embryos. Loading cells with 1 mM BAPTA significantly decreased the number of motoneurons present in the spinal cord at 24 h, indicating that calcium signals are important for normal motoneuron differentiation. Furthermore, in those BAPTA-filled cells that did adopt a motoneuron cell fate, axogenesis was found to be inhibited, suggestive of a role for calcium signalling in neurite initiation. This work provides evidence that calcium signals are necessary at several stages of motoneuron development in vivo.

A narrow window of intracellular calcium concentration is optimal for neurite outgrowth in rat sensory neurones.

We have examined neurite outgrowth in rat sensory neurones when cytosolic free calcium concentration ([Ca2+]i) was varied in the range 0-60 nM. Neurite outgrowth was maximal at 35 nM [Ca2+]i and was reduced at higher and lower values of [Ca2+]i. These results provide direct evidence for Mattson and Kater's suggestion of an optimal calcium range for growth cone function.

Sensitization in voltage clamped barnacle photoreceptors.

1. The light response of Balanus lateral eye photoreceptor was studied using intracellular recording and voltage clamp techniques. 2. The effect upon the light response of a prior depolarization was studied. Two effects were observed: a transient sensitization which appeared not to involve a change in time course, and a long-lived desensitization which was accompanied by a reduction in response time course. 3., The sensitizing effect of depolarizing prepulses increased monotonically with prepulse duration; it was maximal for pulses to around +20 mV. 4. Sensitization was inhibited by perfusion with salines containing cobaltous ions or with calcium-free saline. 5. It is concluded that while desensitization is an effect upon the transduction mechanism, sensitization is a direct influence of a prior depolarization on the light-activated channels.

Immediate and neurotoxic effects of HIV protein gp120 act through CXCR4 receptor.

Primary rat hippocampal neurones show pronounced elevations of intracellular calcium within minutes of exposure to the HIV coat protein gp120. Culture of hippocampal neurones with gp120 causes significant neurotoxicity. We find that the peptide VSLSYRCPCRFF, a competitive inhibitor of the CXCR4 chemokine receptor, markedly inhibits toxicity and eliminates the acute calcium elevation. CXCR4 receptors are thought to signal to the Gi/Go family of trimeric GTP binding proteins. Pretreatment of hippocampal neurones with pertussis toxin to inactivate Gi/Go proteins markedly reduced gp120 neurotoxicity. These results indicate that both short and long term effects of gp120 are the result of activation of the CXCR4 receptor.

Measurements of calcium transients in the soma, neurite, and growth cone of single cultured neurons.

Voltage-gated changes in cytosolic free calcium ion concentration were measured in single, differentiated cells of mouse neuroblastoma clone N1E-115 using the calcium-sensitive dye arsenazo III (AIII). In cells bathed in normal medium containing 10 mM calcium, the changes in AIII absorbance during a single action potential indicated an increase of 1.4 nM in cytosolic calcium. When 10 mM tetraethylammonium (TEA) was added to the bath, the action potential became prolonged and the change in cytosolic calcium increased to 3.9 nM. Under these conditions, repetitive stimulation at 0.5 Hz or faster caused a gradual decline in the amplitude and duration of the action potential and a gradual decline of the change in cytosolic calcium associated with each action potential. The amplitude of the prolonged after-hyperpolarization (AHP) that follows the action potential was found to reflect the magnitude of the change in cytosolic calcium. An action potential elicited in the cell soma caused an increase in cytosolic calcium in the soma, neurite, and growth cone regions of a single cell, indicating that the membrane of all three regions possesses voltage-gated calcium channels. Estimation of calcium flux per unit area of membrane suggests a distinct topographical organization of calcium channels. Calcium channel densities in the growth cone and cell soma regions are similar and significantly higher than that in the neurite.

Calcium channels in neuroblastoma cell growth cones.

The concentration of free calcium ions in the cytosol has been shown to influence many components of growth cone behaviour, including the extension of filopodia and veils, the addition of new membrane to the plasmalemma, the retraction and disappearance of filopodia, and gross collapse and retraction of the growth cone. A spatially localized modulation of these processes by very local calcium changes has been proposed to underlie the steering of growth cones by gradients of neurotransmitters, voltage and cell adhesion molecules. Such local control can be studied in mouse neuroblastoma cells, where depolarization causes calcium to rise in a limited number of spatially restricted hotspots, triggering a localized advance. We have studied the simple, club-shaped growth cones that are characteristically found on advancing neurites. Depolarization caused calcium to increase most at the distal, leading tip. Agents that disrupt calcium-induced calcium release do not affect growth cone calcium dynamics, ruling out a local release of calcium at the tip as a cause of the gradient. Using cell-attached patch recording, we find that L-type calcium channels are present at a higher density at the distal tip than in the proximal growth cone. Our results show that the calcium gradients seen in depolarized growth cones are a direct consequence of a gradient of calcium channel density.

Use of fluorescently labelled calmodulins as tools to measure subcellular calmodulin activation in living dorsal root ganglion cells.

We have used fluorescently labelled calmodulins to probe the activity of calmodulin in living dorsal root ganglion cells. Calmodulin labelled with the fluorophore 5-([4,6 dichlorotriazin-2yl]amino)-fluorescein (FL-CaM) does not change its fluorescence when it binds calcium, while calmodulin labelled at lysine 75 with 2-chloro-(6-(4-N,N-diethylamino-phenyl)-1,4,5-triazin-4-yl (TA-CaM), an environment-sensitive probe, increases its fluorescence when it binds calcium. We micro-injected FL-CaM or TA-CaM into rat dorsal root ganglion cells and found that both probes localise to the cell nucleus. In contrast, endogenous cellular calmodulin, in dorsal root ganglion cells as in hippocampal neurones, is predominantly cytosolic unless the neurones are depolarised, then it moves to the nucleus. FL-CaM and TA-CaM, introduced into dorsal root ganglion cells via a patch pipette, also immediately move to the nucleus, indicating that the nuclear localisation is a property of the labelled calmodulins. Although the subcellular distribution of FL-CaM and TA-CaM does not necessarily match that of endogenous calmodulin, we show that FL-CaM can be used as a control for TA-CaM when studying calmodulin activation in different cellular compartments.

Injection of guanosine and adenosine nucleotides into Limulus ventral photoreceptor cells.

1. Several nucleotide and nucleotide analogues had striking effects when pressure-injected into Limulus ventral photoreceptor cells. The poorly hydrolysable GTP analogues guanosine 5'-0-(3-thiotriphosphate) (GTPgammaS), guanylyl imidodiphosphate (Gpp[NH]p) and guanylyl (beta, gamma methylene) diphosphonate (Gpp[CH(2)]p) produced large increases in the frequency of ;discrete events' that were recorded from photoreceptors in darkness. This effect was only observed after the injected cell was exposed to light. Injection of the ATP analogue ATPgammaS had effects similar to those of the GTP analogues.2. We conclude that GTPgammaS, Gpp[NH]p, Gpp[CH(2)]p and ATPgammaS act at a common site to cause a light-dependent, long-term activation of the excitation mechanism of the photoreceptor.3. Injection of GTP or GDP at pH 4.8 was followed by a smooth, transient depolarization that was observed neither when GTP at pH 7.5 was injected nor when ATP, 5'GMP or 2-[N-morpholino] ethane sulphonic acid (MES) were injected at pH 4.8. The reversal potential of the current induced by GTP injection was significantly more positive than the reversal potential of the light-induced current.4. We conclude that GTP injection induces changes of membrane conductance either in addition to, or different from, the light-induced change of membrane conductance.5. Injection of the ATP analogue adenylyl imidodiphosphate (App[NH]p), and the pyrophosphate analogue imidodiphosphate (p[NH]p) produced a drastic decrease in the sensitivity of photoreceptors to light. This decrease in sensitivity was partially reversed when the concentration of calcium ions in the bathing medium was reduced.6. We suggest that App[NH]p and p[NH]p injections act by increasing the cytoplasmic concentration of calcium ions.

Light adaptation of invertebrate photoreceptors: influence of intracellular pH buffering capacity.

1. The possible role of pH changes in mediating light adaptation in Limulus ventral photoreceptor cells was studied by intracellular injection of zwitterionic pH buffers. The intracellular concentration of buffer was estimated by inclusion of a radioactive marker in the injection solution. 2. The light-induced increase of intracellular Ca2+ concentration was monitored by intracellular aequorin. The light-induced increase of Ca2+ concentration was not markedly altered by injection of pH buffer to an intracellular concentration of about 200 mM. 3. The progressive decrease in responsiveness during intracellular ionophoretic injection of Ca2+ was not markedly altered by injection of pH buffer to an intracellular concentration of about 200 mM. 4. Photoreceptors of both Limulus and Balanus were impaled with two micropipettes and voltage clamped. Membrane current induced by a prolonged steady illumination declined from an early transient to a plateau. This delayed decline of current indicates a light-induced reduction of sensitivity (i.e. light adaptation). The wave forms were similar before and after injection of pH buffer to an intracellular concentration of about 200 mM. 5. We conclude that it is unlikely that a light-induced change of cytosolic pH mediates light adaptation in Limulus (and Balanus) photoreceptors.

Expression of T-type calcium current precedes neurite extension in neuroblastoma cells.

1. N1E-115 mouse neuroblastoma cells morphologically differentiate by extending neurites in a period of seven days after addition of 2% DMSO to the culture medium. We used the whole-cell patch clamp technique to measure calcium currents in these cells under conditions where voltage clamp of the whole membrane was assured. 2. Current densities of both T and L type calcium currents were identical in cells included to differentiate with dibutyryl cyclic AMP and cells induced to differentiate with dimethylsulphoxide (DMSO). Cells differentiated with DMSO were used for all subsequent experiments. 3. All morphologically differentiated cells showed a T type calcium current. In contrast, a minority of morphologically undifferentiated cells did not show a T current. 4. Once expressed, both T and L currents did not change either in current density or in behaviour over a period of five days. 5. These data demonstrate that expression of a T current always precedes neurite extension, and suggest a role for calcium currents in triggering morphological differentiation.

Calcium ion, an intracellular messenger of light adaptation, also participates in excitation of Limulus photoreceptors.

Photoreceptor cells of Limulus ventral eyes were bathed in artificial sea water (ASW) that contained 10 mM-EGTA and no added Ca2+ (EGTA-ASW). Test flashes elicited responses that increased to a maximum size within 10 min in EGTA-ASW but did not change further when dark-adapted cells were bathed for an additional 35 min in this solution. Light responses progressively declined from this maximum size if the cells were repetitively illuminated in EGTA-ASW. In this state of reduced responsiveness, response amplitudes were further reduced by intracellular ionophoretic injection of EGTA; response amplitudes were increased by intracellular ionophoretic injection of Ca2+. Both of these findings are opposite to what is normally observed for cells bathed in ASW. Also, after repetitive illumination in EGTA-ASW, both the slope of the response versus intensity relationship became steeper and light responses often had a delayed increase in amplitude. The light responses and the response versus intensity relation returned to normal when the bathing medium was changed back to ASW containing 10 mM-Ca2+. The light-induced rise in luminescence recorded from photoreceptors injected with the photoprotein aequorin (the 'aequorin response') declined by at most 50% after dark-adapted photoreceptors were bathed in EGTA-ASW for 45 min. However, the aequorin response progressively declined by 98% if cells were repetitively illuminated while bathed in EGTA-ASW. The total intracellular Ca content of whole end-organs was measured by atomic absorption spectroscopy. Total intracellular Ca content did not change significantly while photoreceptors were bathed in EGTA-ASW even after repetitive illumination. We suggest that cytosolic Ca2+ is required by one or more steps in the mechanisms that link rhodopsin isomerization to both (i) an increase in the conductance of the cell membrane to Na+ and (ii) a release of Ca2+ from a light-labile store.